SUMMARY
As a result of this introduction to an anatomical study of cuckoos and touracos, I agree with those who have proposed that the touracos (Musophagidae) deserve ordinal rank, the Musophagiformes. Table 4 presents comparative data on the cuckoos and on Tauraco leucotis donaldsoni. Because they are present in all genera studied for this paper, I have not placed emphasis on two muscles that I recently (1959) proposed be added to muscle formulae: "F," M. plantaris; "G," M. popliteus. They are included in Table 4. The following wing muscles are absent in Tauraco leucotis donaldsoni: cucullaris, pars propatagialis; biceps slip; anconaeus coracoideus; entepicondyloulnaris; flexor metacarpi brevis. The following leg muscles are absent in Tauraco: iliotrochantericus medius ("C"); iliacus ("E"); extensor brevis digiti III; adductor digiti II; adductor digiti IV. The relatively large assemblage of zygodactyl birds that have been grouped together as cuckoos contains genera exhibiting a considerable diversity in internal morphology. It is an old group, first reported from the Oligocene Epoch. One might speculate that the ancestral cuckoos had a more complete complex of thigh muscles than do the modern forms, that they had a tracheobronchial syrinx, and that they had a propensity for developing parasitic nesting habits. Two lines of specialization might then be proposed. One group, with primarily terrestrial habits, tended to retain the thigh muscles but developed a bronchial syrinx. The other, more arboreal, group retained the tracheobronchial syrinx, tended to lose certain thigh muscles, and tended to develop parasitic breeding habits. However, there seems to be little correlation between the type of sternum and the other anatomical characters studied. Some of the cursorial genera (e.g., Geococcyx) have a double-notched sternum, whereas others (e.g., Carpococcyx) have a single-notched sternum. Similarly, both types of sternum are found among the more arboreal cuckoos. In view of the presumed occurrence of the bronchial type of syrinx in some, but not all, owls (as well as in a few other birds), one might question the significance of the bronchial syrinx, especially because of the several patterns found among the cuckoos. Data presented in this paper reveal some of the pitfalls encountered when one compares only a few anatomical characters in single representatives of two or more families of birds. To illustrate that it is unwise to draw broad conclusions on relationships after dissecting only one or two representatives of each of several families, we may cite as examples the relative development of M. iliotibialis in Tauraco and in Cuculus and the intergeneric differences in relative development of M. peroneus longus or M. pronator superficialis and M. pronator profundus among the cuckoos.
SOME ANATOMICAL CHARACTERS OF THE CUCULIDAE
AND THE MUSOPHAGIDAE
BY ANDREW J. BERGER !
OST authors have placed the African touracos ("plantain-eaters") and
the cosmopolitan cuckoos in a single order, the Cuculiformes or Cuculi
(e.g., Mayr and Amadon, 1951, Wetmore, 1951). Bannerman (1933),
Moreau (1938, 1958), Lowe (1943), and Verheyen (1956a, 1956b), however,
believed that the touracos deserve ordinal rank, the Musophagiformes. I agree
with these authors but not for some of the reasons they cite.
I have been interested in the anatomy and relationships of these two groups
of birds for over a decade, but it now seems unlikely that it will be feasible to
continue this work. Consequently, I have decided to publish certain informa-
tion which, akhough still very incomplete, may prove useful to other investi-
gators. The complete appendicular myology of a touraco apparently has never
been described. My plan, therefore, is to describe this musculature for
Tauraco leucotis donaldsoni ("Turacus donaldsoni") and then to compare the
royology, osteology, and pterylosis of this touraco with that of the cuckoos.
Such a comparison can not be made properly without also analyzing differ-
ences in morphology within the family Cuculidae.
This paper is based on the dissection of one or more specimens of the follow-
ing genera and species of cuckoos: Carpococcyx radiceus, Centropus bengal-
ensis, C. superciliosus, Ceuthmochares aereus, Chrysococcyx cupreus, Ch.
("Lampromorpha") klaas and caprius, Ch. ("Chalcites") basalis, Clamator
jacobinus, Coccyzus americanus, C. erythropthalmus, Coua serriana, C.
reynaudii, C. ru]iceps, C. cristata, C. caerulea, Crotophaga sulcirostris, Cuculus
canorus and sp., Dromococcyx pavoninus, Geococcyx cali]ornianus, Guira
guira, Morococcyx erythropygus, Piaya cayana, Phaenicophaeus pyrrhoce-
phalus, Saurothera merlini, Surniculus lugubris, Tapera naevia. These 18
genera represent less than half of the 38 genera recognized by Peters (1940),
although several of his genera have been suppressed by later workers (e.g.,
Mayr, 1944; Delacour and Mayr, 1945; Delacour, 1946; Berger, 1955b).
Consequently, it must be borne in mind that conclusions and generalizations
presented here may have to be altered when data become available on the
remaining genera of cuckoos. As a matter of convenience for other workers, I
have, by and large, used Peters' genera and species of cuckoos. I do this in
part because I do not agree with those who propose that it is not necessary to
examine the internal anatomy of a given species merely because an earlier
genus has been suppressed and its contained species have been placed in some
other genus. Evidence presented here certainly casts doubt on the wisdom of
I dedicate this paper to Erwin Stresemann on the occasion of his 70th birthday.
60
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 61
Berger
synonymizing some genera of cuckoos, as has been done on the basis of
external characters alone.
Over the years I have received alcoholic specimens from several generous cooperators:
Dean Amadon, Jacques Berlioz, Ernest P. Edwards, Herbert Friedmann, Mrs. Richard R.
Graber, Theodore Miller, Renaud Paulian, Helmut Sick, George M. Sutton, P. B. Uzzell,
and John G. Williams.
Some of the material from museums had been collected many years ago. In
a few specimens, an incision through skin and muscle had been made from
chin to vent. It is difficuk (and sometimes impossible) to determine accurately
the details of the ventral feather tracts in such birds. In two specimens, the
articulations of the sternal ribs had been cut bilaterally and the viscera were
held in place by strings tied around the body. When the strings were cut, the
sternum fell out of the body! Shot damage to the specimen of Dromococcyx
was so great that I could not study some important anatomical details. I made
a plea several years ago (1955a) for better anatomical material. For the
benefit of all students of pterylosis, myology, etc., I repeat that plea here. One
often feels that it is better to have no representative of a species or genus than
to have a specimen so mutilated by shot damage or by incisions that one can
obtain only a portion of the information that one needs.
Anatomists sometimes are criticized for publishing information based on
the dissection of only one or two specimens of a species. To be sure, the
anatomist would be delighted to have a dozen or more specimens of each form
that he is studying, but he is rarely so fortunate. In order to obtain the speci-
mens that form the basis for this paper, I wrote well over a hundred letters to
all parts of the world during a 12-year period. Perhaps this is the best
example that I could cite to emphasize the continuing great need for well-
preserved anatomical materiM.
WING MUSCLES OF TAURACO LEUCOTIS DONALDSONI
M. latissimus dorsi. Pars anterior arises from the neural spines of the last two cervical
and the first dorsal vertebrae. It inserts fleshily over an area 7 ram. long beginning 9 min.
from the proximal end of the head of the humerus. Pars posterior arises from the neural
spines of the five dorsal vertebrae. It inserts by a flat tendon (2 min. wide) immediately
posterior to the proximal end of the insertion of pars anterior. Pars metapatagialis, the
dermal component, is a very small band of fleshy fibers about 1 mm. in width. It arises
from the superficial surface of pars posterior at the level of the spine of dorsal vertebra
number 4; it inserts into the skin at the posterior margin of the humeral feather tract
adjacent to the insertion of M. serratus metapatagialis.
M. rhomboideus super/icialis arises by an aponeurosis from the neural spines of the last
three cervical and the first two dorsal vertebrae. The fleshy fibers pass outward and
forward to insert on the anterior 36 mm. of the scapula.
M. rhomboideus pro/undus arises by an aponeurosis from the neural spines of the last
two cervical and the first two dorsal vertebrae. The muscle inserts on the posterior 28
ram. of the scapula.
March 1960
62 THE WILSON BULLETIN Vol. 7, No. I
M. cucullaris, pars propatagialis.I did not find any tendon of this muscle inserting on
the tendon of the tensor patagii longus.
M. serratus pro/undus has a typical origin from the posterior cervical vertebrae and
from the cervicodorsal ribs. The several fasciculi pass backward to insert on the medial
sttrface of the scapula.
M. serratus anterior arises by a superficial and a deep band of fibers. The larger and
more superficial slip arises from the lateral surface of the first true rib, ventral to the
uncinate process. The deeper slip arises from the ventral portion of the last cervicodorsal
rib. The two slips fuse and give rise to a strong aponeurosis (6 mm. wide), which passes
upward between the two heads of M. subscapularis to insert on the ventral edge of the
scapula just caudal to the glenold fossa.
M. serratus posterior arises by fleshy fasciculi primarily from the uncinate processes of
true ribs numbers 1, 2, and 3. The complex is an extensive, but thin, sheet of fleshy fibers
that inserts on the apex of the scapula (by fleshy fibers) and on the ventral edge of that
bone (by an aponeurosis) as far craniad as the insertion of M. serratus anterior. Pars
metapatagials, the dermal component, has a typical origin. The belly, about twice as
large as pars metapatagialis of M. latissimus dorsi, passes upward to insert into the skin at
the posterior margin of the humeral feather tract.
M. proscapulohumeralis is a thin strap of fleshy fibers 4 min. wide at its origin from the
scapula immediately posterior to the glenold lip and the origin of M. scapulotriceps. The
belly is about 13 mm. in length. It has a typical insertion on the humerus between the two
heads of origin of M. humerotriceps.
M. dorsalis scapulae arises from the lateral surface of the scapula in its posterior 33 mm.
It inserts on the anconal surface of the bicipital crest of the humerus.
M. subscapu[aris has the usual two heads.--Pars externa arises for a distance of 8 mm.
from the ventrolateral edge of the scapula, beginning a short distance caudal to the glenold
fossa. Pars interms arises from an area about 20 mm. long on the medial surface of the
scapula. The two heads fuse and insert by a short stout tendon on the internal tuberosity
of the humerus, adjacent to the insertion of M. subcoracoideus.
M. subcoracoideus arises by two heads, a clavicular head and a coracoidal head. The
clavicular head is unusually well developed. It has an extensive origin from an area 17
mm. long on the medial surface of the superior end of the clavicle and from the acromion
of the scapula. The coracoidal head arises from the basal 10 mm. of the posterior face
of the coracold and from the adjacent coracoclavicular membrane. The two heads fuse and
insert in common on the internal tuberosity of the humerus.
M. coracobrachialis anterior is a well-developed muscle, although it does not cover the
anterior surface of the head of the humerus. It arises from the head of the coracold (and
adjacent ligaments) immediately dorsal to the origin of M. biceps brachii. The belly
(about 13 mm. long and 5 mm. in maximum width) passes outward, closely applied to the
capsule of the shoulder joint, to insert on the proximal end of the humerus.
M. coracobrachia[is posterior has a typical origin and insertion.
M. supra'coracoideus has a typical origin from the sternum and the coracoclavicular
membrane. A very large tendon of insertion is formed; the fleshy fibers continue through
the triosseal canal (see p. 82) with the tendon and cover it almost to the point of
insertion on the humerus.
M. sternocoracoideus has a typical origin and insertion.
M. pectoralis. Pars thoracicus is very well developed; it has a typical origin and inser-
tion. Pars propatagialis longus et brevis is a large fleshy fasciculus arising from the
ventral surface of pars thoracicus. The fleshy bundle fuses with the distal end of the belly
Andrew $. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 63
Berger
of M. tensor patagii longus, but it also gives rise to a flat bandlike tendon which appears
to form the bulk of the tendon of M. tensor patagii brevis. The belly of the latter muscle
appears to insert into the tendon formed from pars propatagialis. Near its formation, the
tendon of pectoralis pars propatagialis has a tendinous anchor to the anterior edge of the
deltoid crest of the humerus.
M. deltoideus major has a total length of belly (from the scapular origin) of about 44
min.;the terminal tendinous insertion reaches to within 12 min. of the distal end of the
humerus. The longer head arises from the acromion of the scapula and by a strong
aponeurosis from the ventral edge of that process, as in Coua caerulea. The much smaller
anterior head arises from the very large os humeroscapulare; it inserts by fleshy fibers on
the ancohal surface of the deltoid crest.
M. deltoideus minor is, relatively, a fairly well-developed muscle in the touraco. It is a
flat band of fleshy fibers 3 ram. wide and 14 min. long. Unlike the origin in Coua, it
arises from the acromion and from the coracoscapular and acromioclavicular ligaments. It
has a typical insertion at the junction of the articular head and the deltoid crest of the
humerus. The belly, which passes anterior to the os humeroscapulare and overlies the
subjacent tendon of M. supracoracoideus, is visible anterior to M. deltoideus major after
removal of Mm. tensores patagii longus et brevis.
Mm. tensores patagii longus et brevis are represented by a common belly which arises
from the dorsal and roedial surfaces of the apex of the clavicle and the tip of the acromion;
posteriorly, the origin is shared with the posterior head of M. deltoideus major. The origin
of this complex in Tauraco leucotis is like that found in the cuckoos and not like that
illustrated for Tauraco corythaix by Lowe (1943). The insertion of the tensor patagii
brevis seems to be similar in the two species, and I suspect that Lowe's separation of the
tendon into two discrete slips in corythaix may have been an artifact. At any rate, in
leucotis the tendon of the brevis is a single sheet (about 1.5 min. wide) which inserts
primarily into the belly of M..extensor metacarpi radialis, but an extension of this tendon-
sheet proximally passes posteriorly over the forearm muscles to fuse with the antebrachial
fascia, which is closely adherent to the bases of the proximal secondaries.
M. biceps brachii has a typical double origin from the coracoid and from the bicipital
crest of the humerus. The belly, 37 min. in length, extends to the distal end of the
humerus. The tendon bifurcates to insert 2 ram. and 3 min., respectively, from the
proximal ends of radius and ulna. A biceps slip is not present.
M. triceps brachii exhibits about the same development as in the cuckoos. The scapulo-
triceps portion has a somewhat two-headed fleshy and tendinous origin: one head from the
lateral surface of the scapula, dorsal to the glenoid fossa; the other head from the ventro-
lateral edge of the scapula just caudal to the glenoid lip. There is no humeral anchor. The
humerotriceps can be divided into two heads only proximally, where the two heads arise on
either side of the insertion of M. proscapulohumeralis. There is a typical insertion on the
olecranon and the adjacent area on the proximal end of the ulna.
M. expansor secundariorum is similar in structure to that previously described for
several cuckoos in that there is both a humeral and a scapular origin. The two tendons of
origin give rise to fleshy fibers that insert on several of the proximal secondaries. The
short tendon arises from the distal end of the humerus. The long tendon extends proximad
through the metapatagial skin fold to the axilla where the tendon bifurcates. The dorsal
branch attaches to the medial surface of the scapula, deep to the origin of M. subcoracoid-
eus and at about the line of separation between this muscle and M. subscapularis (pars
interna). The ventral branch attaches to the coracoid in relation to the origin of M.
coracobrachialis posterior; this attachment I have not found in the cuckoos (see p. 77).
March 1960
64 THE WILSON BULLETIN vo. 72, No. 1
M. anconaeus coracoideus.I did not find this muscle in the specimen I dissected.
M. brachialis is typical in origin and insertion.
M. promztor superlicicdis is a well-developed muscle (belly 35 mm. long) which extends
to within 10 mm. of the distal end of the radius.
M. pronator prolundus, with a belly 32 mm. long, extends distad almost as far as M.
pronator superficialis. The origin of both muscles is typical.
M. Ilexor digitorum superlicialis has a typical origin from the distal end of the humerus.
The tendon of origin expands into a broad aponeurosis (the humerocarpal band) which
extends the entire length of the ulna and fuses with the fascia around the bases of the
secondaries; near the distal end of the ulna, the humerocarpal band gives rise to fleshy
fibers (essentially a second belly for the muscle), which fuse with the main belly just
before the tendon of insertion is formed. The main belly (30 min. long) arises from the
deep surface of the humerocarpal band, beginning 12 ram. from its humera1 attachment.
The muscle inserts on the base of the proximal phalanx of digit I[.
M. [lexor digitorum profundus arises by a single head from the ulna, posterior to the
area of insertion of M. brachialis. The belly is about 33 mm. in length; it extends to the
distal end of the ulna. The main tendon of insertion passes around the pisiform process of
the carpometacarpus and then passes distad superficial to the tendon of the flexor digi-
torum superficialis to insert on an extensive area of the distal phalanx of digit II. A very
unusual feature is the presence of a second tendon given off the main tendon opposite the
base of the pollex; this tendon inserts on the palmar surface of the pollex.
M. flexor carpi ulnaris is a well-developed musele, whose belly (37 min. long) extends
nearly the entire length of the forearm. The origin and insertion are typical, but fleshy
fibers do not insert on the bases of the secondaries as they do in Coua caerulea.
M. extensor metacarpi radialis has a typical origin from the lateral epicondyle of the
humerus. The belly is about 34 mm. long. The tendon of the tensor patagii brevis musele
inserts into the belly and its enveloping fascia about 10 mm. distal to the humerus. The
tendon of M. extensor metacarpi radialis inserts on the extensor process of the carpometa-
carpus. M. abductor pollicis arises from the tendon a short distance before its insertion.
M. flexor metacarpi radialis arises by a tendon attached to the lateral epicondyle of the
humerus. The belly is about 28 mm. long, and the fleshy fibers begin about 10 mm. from
the humera1 origin of the tendon. A fascial extension of the tendon of origin fuses with
the fascia surrounding the bases of the proximal secondaries. The tendon inserts on
metacarpal II, about 6 mm. from the proximal end of that bone at the proximal limit of
the intennetacarpal space.
M. extensor digitorum communis arises by a tendon from the lateral epicondyle of the
humerus. The fleshy fibers of the 33 mm.-long belly begin almost at the humeral origin of
the tendon. Two tendons are formed near the base of the carpometacarpus. The shorter
tendon inserts on the base of the pollex. The longer tendon runs in a groove on the dorsal
surface of metacarpal II; near the distal end of that bone, the tendon turns abruptly
around a bony tubercle to insert on the base of the proximal phalanx of digit II.
M. anconeus has a typical origin from the humerus. The belly (39 ram. long) extends
nearly the entire length of the ulna.
M. supinator has a typical origin from the humerus. The belly is short (about 18 mm.),
extending less than half the length of the radius, on which bone it inserts.
M. extensor indicis longus is a long (26 min.), thin muscle arising from the radius only.
fts tendon inserts on the base of the distal phalanx of digit II.
M. flexor metacarpi brevis is absent bilaterally.
M. flexor carpi ulnaris brevis is a short muscle, with a belly 13 mm. long, arising from
Andrew J.
Berger
ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 65
the ventral surface of the distal third of the ulna. The tendon has a typical insertion on
the dorsal surface of the base of the carpometacarpus.
M. extensor pollicis longus is a well-developed muscle with two heads of origin; the
over-all length of the belly is 34 min. The smaller head (24 nun. long) arises from the
middle half of the radius. The larger head (30 nun. long) arises from the ulna, beginning
at the level of insertion of M. biceps brachii. The tendon inserts on the extensor process
of the carpometacarpus deep to the tendon of M. extensor metacarpi radialis.
M. extensor pollicis brevis is a triangular muscle with two heads of origin and with a
total length of about 6 mn. The larger head arises from the base of the carpometacarpus.
The sxnaller head arises from the base of the extensor process and from the tendon of
insertion of M. extensor pollicis longus.
M. abductor pollicis, wit.h a single belly 6 min. long, has a typical origin from the tendon
of insertion of M. extensor metacarpi radialis. It inserts by fleshy fibers on the pollex.
M. adductor pollicis arises fleshily from the anterior surface of the carpometacarpus
near the base of the extensor process. It inserts by fleshy fibers on the posterior face of
the pollex.
M. ]lexor pollicis is a very small band, less than 1 nun. in width, which arises from the
carpometacarpus; it inserts on the base of the pollex, posterior and adjacent to the inser-
tion of M. abductor pollicis.
M. abductor indicis is a well-developed muscle, arising from an extensive area on the
base of the carpometacarpus, both anterior and posterior to the pisiform process, and from
nearly the entire length of the anterior surface of metacarpal II. The muscle inserts by a
stout tendon on the anterior surface of the proximal phalanx of digit II.
M. ]lexor metacarpi posterior has a typical origin by a strong flat tendon from the distal
end of the ulna. The well-developed belly inserts on the posterior surface of about the
proxixnal half of metacarpal III, and it also sends fleshy fasciculi into the bases of the
proximal five or six primaries.
M. ]lexor digiti HI arises along a narrow line from most of the palmar surface of meta-
carpal III, beginning at the proximal lixnit of the intermetacarpal space, and from the
posterior surface of that saxne bone distal to the insertion of M. flexor metacarpi posterior.
The muscle inserts by a tendon on the base of digit III and by fleshy fibers on the entire
anconal surface of that digit.
Mm. interossei dorsalis et palmaris have typical origins from the facing surfaces of meta-
carpals II and III. M. interosseus dorsalis inserts by a tendon on the base of the distal
phalanx of digit II. M. interosseus palmaris inserts by a tendon a little beyond mid-length
of the distal phalanx of digit II.
"LEG" MUSCLES OF TAURACO LEUCOTIS DONALDSONI
M. sartorius arises from the anterodorsal end of the ilium only, i.e., it has no origin from
the neural spine of t.he last dorsal vertebra. It inserts on the anteromedial corner of the
head of the tibiotarsus.
M. iliotibialis is a very extensive muscle, covering all other muscles on the lateral aspect
of the thigh except for parts of Mm. sartorius and semitendinosus. The central part of
the complex is aponeurotic in its distal three-fifths. The muscle arises from the anterior
iliac process and from the entire length of the anterior and posterior iliac crests. Most of
the origin is by an aponeurosis (including that from the anterior iliac process), but the
origin posteriorly is by fleshy fibers. Anteriorly, the origin is inseparably fused with the
origin of M. sartorius. Distally, the tendon contributes to the formation of the pateliar
March 1960
66 THE WILSON BULLETIN VoL. 72, No. 1
ligament, inserts on the tibiotarsus, and fuses with the deep fascia of the crus, particularly
that covering M. peroneus longus.
M. iliotrochantericus posterior is well developed and has a considerable bulk of its fibers
arising dorsal to the acetabulum. The muscle arises from the anterior iliac process and
from all of the anterior iliac fossa. It inserts on the femur by a strong flat tendon,
beginning 1 mm. distal to the trochanter.
M. iliotrochantericus anterior also is well developed. It arises from the posterior surface
of the anterior iliac process and from the ventrolateral edge of the ilium. It inserts by a
strong tendon (2 mm. wide) on the lateral surface of the femur, beginning 7 mm. distal to
the proximal end of that bone.
M. iliotrochantericus medius is absent bilaterally.
M. glutens medius et minimus is conspicuously developed, especially the long tendon of
insertion, which passes superficial to the tendon of M. iliotrochantericus posterior to insert
on the posterolateral edge of the femur about 7 mm. from the proximal end of that bone
between the areas of insertion of Mm. iliotrochantericus anterior and ischiofemoralis. The
fleshy belly is 5 mm. long; it arises from the most posterior part of the anterior iliac crest.
M. ]emorotibialis externus arises by fleshy and senitendinous bands (especially distally)
from most of the lateral surface of the femur. The origin begins just distal to the level of
insertion of M. iliotrochantericus anterior.
M. /emorotibialis medius arises by tendinous and fleshy fibers from the trochanteric
ridge and the femoral shaft distal to it. The belly is fused with that of M. femorotibialis
externus. Both muscles contribute to the patellar ligament; their configuration is like that
ilhstrated for Co.ua caerulea (Berger, 1953a: Fig. 6).
M./emorotibialis internus is a long thin muscle that arises from the medial surface of
the femur, beginning 14 mm. inferior to the neck of that bone; the belly increases in size
distally. It inserts by a single tendon on the anteromedial surface of the head of the
tibiotarsus.
M. biceps/emoris arises by two heads which fuse at about nid-thigh level. The anterior
head is somewhat larger; it arises by an aponeurosis from the caudal portion of the
anterior iliac crest and by fleshy fibers from the anterior 3 mn. of the posterior iliac crest.
The more rounded posterior head arises by tendinous and fleshy fibers from a pronounced
tubercle on the lateral surface of the ilium about 3 mm. ventral to the projecting posterior
iliac crest. The muscle inserts by a stout, rounded tendon on the fibula about 16 mm.
inferior to the head of that bone.
M. ischio/emoralis has a typical origin from the lateral surface of the ischium, dorsal to
the origins of Mm. seminembranosus and adductor longus et brevis. It inserts by a strong
tendinous band (2.5 mm. wide) beginning 5 mn. inferior to the trochanter and immedi-
ately posterior to the insertion of M. glutens medius et mininus.
M. piriformis.--Both parts of this complex are well developed. Pars caudo[emoralis
arises from the base of the pygostyle by a wide (8 mm.) tendinous band and by fleshy
fibers from the superficial surface of this band. The maximum width of the belly, inferior
to the posterior iliac crest, is 10 mm. Pars iliofemoralis arises by fleshy fibers from the
lateral surface of the ilium, ventral to the origin of Mm. semitendinosus and biceps
femoris. The belly passes downward and forward, superficial to pars caudofemoralis, to
insert on the posterolateral edge of the femur for a distance of 7 mm., beginning 10 mm.
inferior to the trochanter. Pars caudofemoralis inserts by a tendon-sheet (4.5 mm. wide),
beginning 12 mm. inferior to the trochanter and immediately medial to and in contact with
the insertion of pars iliofemoralis.
Mm. semitendinosus and accessorius semitendinosi.--The bulky semitendinosus muscle
Andrew J.
Berger
ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 67
arises from the ventral surface of the projecting posterior iliac crest in its posterior 8 mm.;
there is no origin from the free caudal vertebrae. The raphe which separates this muscle
from the accessory semitendinosus muscle behind the knee broadens into a 2 mm.-wide
tendon, which inserts on the medial surface of the tibiotarsus 2 mm. posterior to the
insertion of M. semimembranosus. The accessorius muscle inserts by fleshy fibers on the
posterolateral edge of the femur in its distal 13 mm.
M. semimembranosus arises by an aponeurosis (shared in part with M. adductor longus
et brevis, pars externa) attached to the lateral surface of the ischium a few millimeters
dorsal to the ischlopubic fenestra. Most of the thin sheetlike belly of the semimembran-
osus muscle is concealed by the overlying semitendinosus muscle. M. semimembranosus
inserts along a curved line (6 mm. long) on the anteromedial edge of the tibiotarsus,
beginning 7 mm. inferior to the proximal end of that bone.
M. iliacus.--I found no evidence of this muscle in either leg.
M. ambiens is a thin strap of muscle, 30 mm. in length; it has a maximum width of 5
mm. near its origin from the pectineal process. It arises mostly by fleshy fibers. The
tendon of insertion, like that in Coua caerulea, enters the patellar ligament and passes
diagonally downward and laterad around the front of the knee. Below the knee the tendon
fans out into an aponeurosis which gives origin to parts of Mm. flexores perforati digiti
II, III, and IV.
M. obturator internus is a well-developed triangular-shaped muscle arising from the
medial surface of the ischium and pubis; it also has a large portion arising in the renal
depression, as in Coua and several other unrelated birds. The large tendon passes out of
the obturator foramen and inserts on the lateral surface of the femur, proximal to the
insertion of M. ischiofemoralis and beginning about 3 mm. inferior to the trochanter. The
tendon of M. gluteus medius et minimus passes superficial to the tendons of insertion of
Mm. obturator internus and externns.
M. obturator externus is a small band of fleshy fibers arising from the dorsal margin of
the obturator foramen. A flat tendon forms and fuses with the proximal side of the tendon
of M. obturator internus.
M. adductor longus et brevis is composed of two distinct parts. Pars externa arises by
an aponeurosis (8.5 mm. wide) from the lateral surface of the ischium, beginning about 2
mm. caudal to the obturator foramen and 4 mm. dorsal to the ischiopubic fenestra. Pars
externa inserts on the femur medial to M. piriformis and beginning at the inferior limit of
the tendon of pars caudofemoralis and extending distad about 23 mm. Pars interna arises
by a dense aponeurosis 10 mm. wide, beginning at the posterior margin of the obturator
foramen and passing candad along the dorsal margin of the ischiopubic fenestra. Fleshy
fibers form at once in the anterior fifth, but posteriorly the aponeurosis extends distad
about 15 mm. before fleshy fibers arise. Pars interna inserts on the femur mostly by an
aponeurosis (by fleshy fibers superiorly), medial to the insertion of pars externa and
extending for a distance of 33 mm.
M. tibialis anterior, as in other birds, has both a fernoral and a tibial head. The fernoral
head arises by a stout tendon from the anterodistal end of the external femoral condyle.
The tibial head arises in common with, but deep to, M. peroneus longus. The two heads
fuse about a third the way down the crus. The over-all length of the belly is 60 mm.,
and it extends distad as far as the ligamentum transversum. A very large tendon of inser-
tion is formed; it passes deep to the ligamenturn transversum and inserts 7 ram. inferior
to the proximal end of the tarsometatarsus. Just before the area of insertion, the tendon
gives off a fascial band that fuses with the fascia covering the muscles of the anterior
surface of the tarsometatarsus.
68 THE WILSON BULLETIN March 1960
Vol. 72, No. 1
M. extensor digitorum longus arises by fleshy fibers from the proximal third of the
anterior surface of the tibiotarsus and from the inner and outer cnemial crests of that
bone. The belly extends for a distance of 60 mm., but the muscle is very small in its distal
half. The tendon passes deep to the ligamenturn transversum and then under a bony
bridge on the anterior surface of the tibiotarsus. Continuing distad, the tendon is held in
place by a strong ligament (and not by a bony bridge) on the proximal end of the
tarsometatarsus. The tendon trifurcates in the distal third of the tarsometatarsus to supply
digits II, III, and IV. The tendon complex to digit III (the largest digit) is unusually
well developed; the small single tendon to digit II exhibits the poorest development.
Attachments of the tendons are made to each of the phalanges of the respective digits.
M. peroneus lo.ngus arises by semitendinous bands, in common with the underlying
muscles, from the outer and inner cnemial crests and the rotular crest of the head of the
tibiotarsus. The belly is about 50 mm. long, and, although it covers all of the other
muscles on the anterolateral surface of the proximal five-sevenths of the crus, it is a rela-
tively thin sheet of muscle which does not extend around to the medial surface of the crus
(as this muscle does in Coua). The tendon bifurcates near the distal end of the tibio-
tarsus. The shorter of the two tendons inserts on the tibial cartilage and associated liga-
ments. The longer tendon passes downward to insert on the tendon of M. flexor perforatus
digiti IIIa short distance inferior to the proximal end of the tarsometatarsus.
M. peroneus brevis arises from the fibula and the tibiotarsus for a distance of about 45
mm., beginning at the level of insertion of M. biceps femoris. It inserts on the postero-
lateral corner of the proximal end of the tarsometatarsus at the base of the hypotarsus.
M. gastrocnemius has the typical three heads. Pars externa arises primarily by a flat
aponeurotic band fused to the lateral surface of the distal arm of the biceps loop. Pars
media arises from the posterior surface of the distal end of the femur, immediately distal
to the area of insertion of M. accessorius. Pars interna arises from the entire medial surface
of the inner cnemial crest of the tibiotarsus and from the fascia covering the anterior and
medial aspects of the knee joint, where this .head has an intimate relationship to the fleshy
insertion of M. sartorius. None of pars interna covers the anterior or lateral portion of the
crus, as it does in Coua caerulea. Pars media and pars interna are separated by the tendon
of insertion of M. semimembranosus. There is also a strong fascial band extending from
M. semitendinosus to both pars media and pars interna. The very strong common tendon
of M. gastrocnemius inserts on the back of the hypotarsus and the posterior ridges of the
tarsometatarsus throughout most of its length.
M. plantaris has a typical origin from the proximal end of the tibiotarsus. The belly is
about 18 mm. in length. The small tendon of insertion expands distally and inserts on
most of the proximal end of the tibial cartilage.
M. p.opliteus is a reasonably well-developed muscle, about 3 mm. wide and 6 mm. long.
Both the origin and the insertion are semitendinous. It arises from the head of the fibula.
The fleshy fibers pass downward and mesiad to the insertion on the tibiotarsus.
M. ]lexor perforatus digiti H is a small muscle with a belly 35 mm. long but only 3 mm.
wide. It arises from the continuation of the ambiens tendon and from the deep surface of
M. flexor perforatus digiti III. The tendon passes through the deep aspect of the lateral
side of the tibial cartilage. The tendon of insertion is perforated by both of the deep flexor
tendons to digit II (flexor digitorum longus and flexor perforans et perforatus digiti II)
and then inserts on both sides of the proximal phalanx of digit II.
M. ]lexor per/oratus digiti IH arises from the femur (in common with Mm. flexor
perforatus digiti IV and flexor hallucis longus) and from a broad aponeurosis which is a
continuation of the ambiens tendon. The belly is 50 mm. in length. The long tendon of
Andrew I. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 69
Berger
M. peroneus longus inserts on the tendon about 8 min. inferior to the proximal end of the
tarsometatarsus. Near the distal end of the tarsometatarsus, the tendon is united by a
vinculum ("V") with the tendon of M. flexor perforans et perforatus digiti III, which lies
deep to the tendon of M. flexor perforatus digiti III. The vinculum arises from the deep
surface of the latter tendon and then passes deep to the subjacent tendon of flexor
perforans et perforatus digiti III, so that the vinculum fuses with the deep side of the
latter tendon. The very large tendon of insertion of M. flexor perforatus digiti III is
perforated by both of the deep flexor tendons to digit III and then inserts on both sides of
the proximal phalanx and on the base of the second phalanx of digit III.
M. ]lexor per/oratus digiti IV arises from the femur (in common with Mm. flexor
perforatus digiti III and flexor hallucis longus) and from the ambiens tendon. The belly
is 40 min. in length. At the base of digit IV, the tendon of flexor perforatus digiti IV
expands into a large mass, which ensheathes the tendon to that digit of M. flexor digitorum
longus. The tendon then sends slips to attach to the proximal ends of phalanges 2, 3, and
4 of digit IV.
M. ]lexor per/orans et per/oratus digiti II appears to arise exclusively from the lateral
femorofibular ligament (i.e., there is no direct origin from bone), immediately distal to
the attachment of the distal arm of the biceps loop. The belly is 22 mm. long. The tendon
inserts on both sides (the roedial slip being the stronger) of the bones at the inter-
phalangeal joint between phalanges 1 and 2 of digit II.
M. ]lexor per/orans et per/oratus digiti HI arises by fleshy and tendinous fibers from the
outer cnemial crest, the third arm of the biceps loop, and from adjacent ligaments on the
anterolateral aspect of the knee, including the pateliar ligament. The belly is about 40
min. long. The tendon perforates the tendon of M. flexor perforatus digiti III and is itself
perforated by the tendon of M. flexor digitorum longus. It inserts on both sides of the
distal end of phalanx 2, digit III.
M. ]lexor digitorum longus arises by two distinct heads from the proximal end of the
tibiotarsus, most of M. popliteus being visible between the two heads. The belly (55 mm.
long) extends to within 15 mm. of the distal end of the tibiotarsus, so that this muscle is
very well developed as compared with M. flexor hallucis longus. The tendon of M. flexor
digitorum longus is the only one that passes through a bony canal in the hypotarsus.
Inferior to that process, the tendon of the flexor hallucis longus passes diagonally down-
ward and superficial to the tendon. The two tendons are connected by a small vinculum in
the distal fourth of the tarsometatarsus. The tendon of the flexor digitorum longus muscle
trifurcates. The primary insertion of each branch is on the ungual phalanx of digits II,
III, and IV. Vinculae pass from the deep surface of the tendons to other phalanges of
these digits.
M. ]lexor hallucis longus is, compared with the flexor digitorum longus, a weakly-
developed muscle with an over-all belly-length of about 30 ram. It arises by two distinct
but short heads from the distal end of the femur; both heads lie roedial to the tendon of
M. biceps femoris. The more lateral (anterior of Hudson, 1937:49) and fleshy head
arises from the posterior face of the external condyle. The larger medial head arises by
tendinous fibers from the intercondyloid area of the femur; this origin is shared with Mm.
flexores perforati digiti III and IV. The tendon of insertion does not pass through a bony
canal in the hypotarsus, but passes along the lateral surface of that process, where it is
held in place by a fascial band. The tendon inserts on the ungual phalanx of the hallux.
M./lexor hallucis brevis is a bulky muscle whose belly (15 min. long) is limited to the
proximal third of the tarsometatarsus. The tendon expands to ensheath the tendon of M.
March 1960
70 THE WILSON BULLETIN Vol. 72, No. 1
flexor hallucis longus and then inserts on both sides of the base of the proximal phalanx
of the hallux.
M. extensor hallucis longus arises by two heads, one on each side of the tendon of
insertion of M. tibialis anterior. The medial head is the larger of the two. The origin
begins at the proximal end of the tarsometatarsus. The over-all length of the belly is 15
min. The tendon is held in place by fibrous bands at the distal end of the tarsometatarsus.
The single tendon inserts on the base of the ungual phalanx of the .hallux.
M. extenaor propflus digiti III is a long, thin and weakly-developed muscle, whose belly
begins at the proximal end of the tarsometatarsus between M. extensor brevis digiti IV
and the tendon of M. tibialis anterior. The tendon forms at the junction of the proximal
and middle thirds of the tarsometatarsus. Fleshy fibers insert on this tendon to the level
of the trothlea for digit III, so that the over-all length of the fleshy belly is 35 mm. The
tendon expands to insert on most of the dorsal surface of the proximal phalanx of digit III.
M. extensor brevis digiti IV is another long, thin muscle, whose origin begins at the
proximal end of the tarsometatarsus and extends distad about 32 mm. The tendon passes
around the trothlea for digit IV and inserts on the medial side of the base of the proximal
phalanx of that digit.
M. abductor d.igiti II is a fairly stout muscle, 10 ram. long, which arises from the distal
end of the tarsometatarsus. It inserts by a tendon on the dorsomedial corner of the base of
the proximal phalanx of digit II. The tendon is visible throughout its course.
M. lumbrlcalis is vestigial. A few fleshy fibers (about 7 mm. in length) arise from the
deep side of the tendon of M. flexor digitorum longus near the distal end of the tarso-
metatarsus.
M. abductor digiti IV has a small belly 10 mm. long, which is limited to the proximal
end of the tarsometatarsus. The long tendon is held in place by a ligament at the distal
end of that bone. The tendon inserts on the ventrolateral corner of the base of the
proximal phalanx.
COMPARATIVE MYOLOGY
My reasoning in the analysis that follows may be clearer if prefaced by
some general remarks. If we assume with Hudson (1937:77) that the general-
ized condition is represented in those birds that possess the full complement of
formulae and other muscles (as in the Galliformes), it follows that the absence
of muscles represents a specialized condition. We assume also that, by and
large, closely related birds will have lost the same muscles, and that either the
hypertrophy or the loss of leg muscles must in general be correlated with loco-
motor habits or feeding habits (as in the birds of prey, with their very large
flexor digitorum longus and flexor hallucis longus muscles). Further, we
recognize that the age of a genus (and/or its geographical isolation and
related factors) may be such that differences in locomotor pattern within the
genus may not yet be. accompanied by extensive myological changes. Finally,
we assume that what we refer to as a "vestigial muscle" is a muscle that is in
the process of being lost phylogenetically, that the degree of degeneration of
such a muscle may vary among individual specimens of a genus or a species,
and that one might expect some individual variation in the presence or
absence of such a vestigial muscle.
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 71
Berger
Lowe (1943:512-514) made a considerable issue out of the alleged differ-
ence in development of M. iliotibialis between Tauraco corythaix and Cuculus
canorus. I have been puzzled by the fact that I could not in any sense cor-
roborate Lowe's description and illustration of M. iliotibialis in Cuculus.
Actually, there is little resemblance between Lowe's figure of this muscle in
Cuculus and what I found in the specimens available for dissection. Except
for the fact that the central portion of this muscle is not aponeurotic from
origin to insertion, M. iliotibialis in Cuculus is very similar to that muscle in
Chrysococcyx ("Lampromorpha") caprius (Berger, 1955b: Fig. 71), and it
is almost identical to the muscle in Ch. cupreus (Fig. 1).
I1. tib.
Bic. fern.
! cm.
I I
Fit. 1. Lateral view of thigh of Chrysococcyx cupreus to illustrate the relative weak
development of M. iliotibialis. Bic. fem., biceps femoris; I1. tib., iliotibialis; Sar., sar-
torius; Semim., semimembranosus; Semit., semitendinosus.
In Cuculus, M. iliotibialis arises by fleshy fibers from the tip of the anterior
iliac process and by an aponeurosis from the anterior iliac crest and the
cephalic half of the posterior iliac crest. The caudal part of this aponeurosis
March 1960
72 THE WILSON BULLETIN Vol. 72, No. 1
overlies the origin of the anterior portion of M. biceps femoris (as in other
cuculine genera), but the iliotibialis muscle has no direct attachment to the
belly of the biceps muscle (as illustrated by Lowe). The posterior segment
(the gluteus posterior of Gadow) of the iliotibialis muscle is poorly developed
so that it does not conceal any part of Mm. semitendinosus or semimembran-
osus, and it conceals only a small proximal and anterior part of M. biceps
femoris. The central portion of the iliotibialis muscle is aponeurotic in about
its distal half, and few, if any, of the fleshy fibers of the relatively small
anterior and posterior fleshy parts of the complex extend distad as far as the
knee. Relative to other cuckoos, M. iliotibialis is poorly developed in Cuculus
and in Chrysoco.ccyx. In many cuckoos, however, the development of M. ilio-
tibialis is equal to that found in Tauraco. Hence, this muscle does not offer
any clues concerning the relationship between the touracos and the cuckoos.
It might be well to add that, anatomically, Cuculus is perhaps the least
"typical" genus of the family, and, therefore, that it is a poor genus to use as
a basis for comparison with genera of other families. Although much informa-
tion is needed on the habits of other parasitic cuckoos, it would appear that
Cuculus is the most highly specialized genus in the family.
Hudson (1937:14) reported that among American genera of birds, M.
iliacus is "present in all forms examined and is one of the most uniformly
developed thigh muscles in birds." This muscle is present in some genera of
cuckoos, absent in other (Old World) genera: Carpococcyx, Centropus,
Coua, Chrysococcyx, and Cuculus. Moreover, I did not find this muscle in
the specimen of Tauraco. leucotis that I dissected.
M. iliacus is a small muscle at best (with a maximum width of about 1 mm.
in most cuckoos examined), and it might easily be destroyed by shot or be
torn away in dissecting. Except for Carpococcyx, however, I dissected two or
more specimens of each of the other genera of cuckoos mentioned above and
found no evidence of the muscle; it was absent in each of five species of
Coua. Shot damage to both femora of the single specimen of Dromococcyx
available for dissection made it impossible to determine whether or not M.
iliacus is present in that genus. Because this muscle has been found in all
other genera of American cuckoos examined, I assume tentatively that it is
present in Dromococcyx. Relative to other muscles in the thigh, M. iliacus
might well be considered vestigial or rudimentary (with a very weak action)
in many birds. It is highly desirable for someone to conduct a study directed
to determine whether or not the absence of this muscle within a genus or
species is a matter of individual variation. Gadow and Selenka (1891:145)
reported that M. iliacus was absent as an individual variation in Bucorvus
and Platycercus.
Nevertheless, I have proposed (1959) that M. iliacus be included in leg-
Andrew $. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 73
Berger
muscle formulae and that its presence be indicated by adding the letter "E" to
such formulae. I would be among the first to agree that muscle formulae
must be used with caution and that they often are misused. I think that this
is especially the case when contemporary writers ignore data presented by
Hudson and others and include only the symbols proposed many years ago by
Garrod. Thus, if we use only Garrod's symbols, the muscle formula for
Tauraco leucotis is ABXYAm; that for cuckoos is either ABXYAm or
AXYAm. This is misleading because it does not tell the whole story; the
myology of the hind limb is not that similar in the two groups of birds. This
can be demonstrated by using Hudson's expanded muscle formula, with
or without the addition of "E." The formula for Tauraco then becomes
ABDXYAmV. For cuckoos the formulae are:
AXYAm--Cuculus, Chrysococcyx
AEXYAm--Coccyzus, Piaya, Saurothera, Tapera, Clamator, Surniculus
ABXYAm--Coua, Carpococcyx, Centropus
ABEXYAm--Geococcyx, Morococcyx, Dromococcyx (E?), Crotophaga,
Guira, Ceuthmochares, Phaenicophaeus
It is my opinion that the presence of M. gluteus medius et minimus (D)
and of the vinculum (V; between the tendons of Mm. flexor perforatus digiti
III and flexor perforans et perforatus digiti III) in Tauraco and the absence
of these structures in the cuckoos constitute a significant difference in their
pelvic musculature. Another interesting difference is that in the cuckoos the
two deep plantar tendons (the tendons of Mm. flexor digitorum longus and
flexor hallucis longus) pass through separate bony canals in the hypotarsus,
whereas in Tauraco there is a single bony canal that transmits only the tendon
of M. flexor digitorum longus. M. flexor hallucis longus in Tauraco arises
from the femur by two separate heads, both of which lie roedial to the tendon
of M. biceps femoris; in the cuckoos (but apparently not in Centropus) the
muscle arises by a single head from the intercondyloid area of the femur.
Beddard (1898a:45) said that W. A. Forbes found the flexor hallucis
longus muscle to be "totally absent" in Centropus (see also Verheyen, 1956a:
8), but I found it in both species examined. Beddard (1898b:276) further
stated that the deep plantar tendons of Centropus are peculiar in that "no
branch is sent to the hallux," and I found this to be true. M. flexor digitorum
longus in Centropus is at least twice as bulky as M. flexor hallucis longus,
although the latter muscle extends distad about half the length of the tibio-
tarsus. The tendons of both muscles pass through bony canals in the hypo-
tarsus, below which the tendon of M. flexor hallucis longus fuses completely
with the tendon of M. flexor digitorum longus. The resulting common tendon
trifurcates to send branches to insert on digits II, III, and IV; no branch is
sent to the hallux. There exists, however, a most unusual condition that I
74 THE WILSON BULLETIN March 1960
Vol. 72, No. 1
have not seen in any other genus. There is a very large automatic flexor of
the hallux (Berger, 1953a:75). From its attachment to the trochlea and the
fibrocartilaginous pulley for digit IV, the automatic flexor passes mesiad to
the base of the hallux and then runs distad (held close to the bone by a vaginal
sheath) to insert on the entire plantar surface of the base of the distal phalanx
of the hallux..Consequently, the relationships of the automatic flexor to the
phalanges in Centropus are identical to those of the flexor hallucis longus
tendon when it is present as in the other cuckoos. The absence of the tendon of
M. flexor hallucis longus to the hallux in Centropus is compensated for, in
part, because M. extensor hallucis longus not only has a typical insertion on
the dorsal surface of the distal phalanx but it also sends a strong slip to insert
on the roedial and plantar edge of the proximal phalanx of the hallux (for
other examples of this double insertion, see Hudson, 1937:52).
The absence of pars iliofemoralis (B) of M. piriformis (=M. caudofemor-
alis) in some of the cuckoos raises interesting questions. The muscle is absent
in the more arboreal members of the group; it is present in the more terrestrial
members (but also in Eudynamis). In locomotor habits, Guira and Croto-
phaga are intermediate between the cursorial Geococcyx and the arboreal
Coccyzus. Both Guira and Crotophaga possess the full cuculine complement
of leg muscles, and they also share similar social nesting habits. In this
instance, therefore, the similarity in development of leg muscles would seem
to indicate closeness of relationship rather than locomotor pattern. I assume
that the primitive cuckoos possessed both parts of the piriformis complex and
that the loss of pars iliofemoralis in certain modern cuckoos is a secondary
condition correlated with locomotor habits, in this instance with perching and
arboreal habits (often also with parasitic breeding habits). One disadvantage
of muscle formulae, however, is that they indicate nothing about the relative
development of the muscles concerned. It is interesting to note, for example,
that in Ceuthmochares (unlike any of the other genera possessing the muscle)
pars iliofemoralis is vestigial; it is a minute band of fleshy fibers about 1 min.
in width. One might expect this muscle to be absent, either unilaterally or
bilaterally, in some specimens of Ceuthmochares. As a possible similar
example, I have pointed out elsewhere (1956:436) that pars caudofemoralis
of the piriformis complex may be considered to be vestigial in the Sandhill
Crane (Grus canadensis tabida), and Fisher and Goodman (1955:86)
reported that pars caudofemoralis was absent in one of three specimens of the
Whooping Crane (G. americana). In the cuckoos, pars caudofemoralis is
relatively well developed in all species examined, although it is more strongly
developed in the arboreal species. (For a discussion of the function of pars
caudofemoralis, see Fisher, 1957.)
The relative development of Min. peroneus longus and peroneus brevis in
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 75
Berger
birds has been discussed by Mitchell and by Hudson (1937). Mitchell (1913:
1068) said: "I find it extremely difficult to associate the conditions of the
peroneal muscles with differences in habit that point directly to functional
adaptation," and that, with very few exceptions, "there is a close conformity
between the condition of the peroneals and what appears to be the most
securely founded systematic divisions. Birds seem to have this or that type of
peroneal muscle, not because they are arboreal or terrestrial, swimmers or
waders, scratchers, predatory or vegetarian, but because it is the type occur-
ring in this or that systematic division."
One might agree in essence with such a generalization, but an analysis of
these two muscles as found in the cuckoos may give one an insight into the
difficulties involved as one attempts to determine which myological characters
indicate closeness of relationship and which are a reflection of locomotor or
other habits. These two factors are, of course, undoubtedly interrelated.
We may ignore for present purposes the fact that either of the peroneal
muscles may be absent in a few non-cuculine genera. Among the cuckoos,
then, we may say that M. peroneus longus exhibits its two extremes of
development. At its highest degree of development, M. peroneus longus
arises from the proximal end of the tibiotarsus, and its belly (or aponeurosis
of origin) conceals all but the most distal parts of Mm. tibialis anterior and
peroneus brevis (Berger, 1953a: Figs. 5, 8). This situation is found, in
general, in terrestrial cuckoos (Geococcyx, Morococcyx, Centropus, Carpo-
coccyx, Coua, Crotophaga, Guira, Tapera, and (?) Dromococcyx), but also
in Tauraco. In a second group of primarily arboreal cuckoos, M. peroneus
longus is a long, thin, and deeply situated muscle, concealing no part of M.
tibialis anterior (Berger, 1952: P1. II). This condition is found in Coccyzus,
Piaya, Saurothera, Ceuthmochares, Clamator, Chrysococcyx, Cuculus, Surnicu-
lus, and Phaenicophaeus. With one exception (Tapera), the fully-developed
peroneus longus muscle is found in those genera that possess both parts (AB)
of M. piriformis. These genera either are essentially cursorial and terrestrial
birds or they are well-defined genera whose distribution or breeding behavior
appear to set them apart from the remaining cuckoos. Among the cuculine
genera considered in this paper, we may conclude that the relative develop-
ment of M. peroneus longus coincides reasonably well both with locomotor
habits and with the two major subgroups of the family. The relative develop-
ment of this muscle, however, is of no value in determining the relationship
between the cuckoos and the touracos (as Lowe proposed, 1943:514).
The relative development of M. extensor proprius digiti 111 seems to be
correlated with locomotor habits. In the more arboreal cuckoos (and in
Tauraco) this muscle arises just distal (just proximal in Tauraco) to the
insertion of M. tibialis anterior, and the belly extends to the distal end of the
March 1960
76 THE WILSON BULLETIN Vol. 72. No. i
tarsometatarsus: Coccyzus, Piaya, Saurothera, Crotophaga, Guira, Ceuthmo-
chares, Surnculus, Clamator, Phaenicophaeus, Chrysococcyx, Cuculus. In a
second group, the entire muscle is limited to the distal half or less of the
tarsometatarsus: Geococcyx, Morococcyx, Dromococcyx, Tapera, Carpo-
coccyx, Coua, Centropus.
M. adductor digiti II is the smallest muscle of the pelvic limb in cuckoos. It
seems proper to refer to it as a vestigial muscle in this group; it was, however,
present (or absent) constantly in all specimens of the several genera examined.
I did not find this muscle at all in Cuculus, Chrysococcyx, Ceuthmochares,
Surniculus, Phaenicophaeus, Centropus, Piaya, or in Tauraco.
Certain tendon relationships in the foot appear to be correlated, in part,
with zygodactylism, but there are curious exceptions. For example, the
tendon of M. flexor perforatus digiti IV does not ensheathe the tendon of M.
flexor digitorum longus in 16 of the genera studied, but it does ensheathe the
longus tendon in Dromococcyx and in Phaenicophaeus, as well as in the semi-
zygodactylous (see Moreau, 1938:666) Tauraco. Phaenicophaeus presents an
intermediate condition in that about 95 per cent of the tendon inserts on the
lateral side of the basal phalanx of digit IV, whereas the remainder of the
tendon inserts on the roedial side of the same phalanx. The tendon of M.
flexor hallucis brevis forms a complete sheath around the tendon of M. flexor
hallucis longus in all of the cuculine genera (as well as in Tauraco and many
other birds) except Centropus, in which genus the flexor hallucis longus
muscle does not insert on the hallux (see p. 73).
Another curious difference is found in the structure of M. femorotibialis
internus in Cuculus and Coua, two distantly related cuckoos. In these two
genera alone, M. femorotibialis internus has two independent heads, each of
which inserts by its own tendon on the. head of the tibiotarsus. The muscle is
single in the other genera, including Tauraco.
Turning our attention now to the muscles of the wing, we find that Min.
tensores patagii longus et brevis in cuckoos (as in Tauraco) tend to be repre-
sented by a single muscle mass that gives rise to two tendons distally. There
is a single fleshy origin from the dorsomedial surface of the furculum, also
from the adjacent acromion of the scapula in some genera. In Chrysococcyx
two bellies are discrete distally but the complex does not differ significantly
from that in other cuckoos. The primary insertion of M. tensor patagii brevis
in all cuculine genera (despite Lowe's statement and illustration) is on the
tendinous surface of M. extensor metacarpi radialis, and part of the brevis
tendon fans out posteriorly over the forearm muscles to fuse with the ante-
brachial fascia; the pattern is the same in Tauraco (see page 63, and
Garrod, 1881, P1. 22). There are secondary attachments of the brevis tendon-
complex as follows: (1) to the lateral supracondylar ridge of the humerus--
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 77
Berger
Geococcyx; (2) to the lateral epicondyle (ectepicondylar process) of the
humerus, superficial or proximal to the origin of M. extensor metacarpi
radialis--Coccyzus; (3) to the tendon of origin of M. extensor digitorum
communisCo.ccyzus, Morococcyx, Dromococcyx, Tapera, Chrysococcyx.
One must conclude that these secondary attachments are not important taxo-
nomically.
M. expansor secundariorum has both a humeral and a scapular tendon of
origin in all of the cuckoos. At the elbow there may be two independent
bellies (as in Coua, Geococcyx, Guira, etc.) or a single muscle mass with two
fleshy tongues directed proximally (as in Crotophaga, Dromococcyx, Surnicu-
lus, Phaenicophaeus) to the two primary (humeral and scapular) tendons of
origin. This muscle is similar in Tauraco, but the axillary tendon bifurcates,
the two resulting branches attaching respectively on the scapula and on the
coracoid, i.e., the pattern of attachment is "ciconine" rather than "cuculine"
(Garrod, Beddard). It seems likely that this is a significant difference between
Tauraco and the cuckoos.
M. flexor metacarpi brevis is a small muscle, arising from the base of the
carpometacarpus or from the os radiale and associated ligaments, and insert-
ing on the tendon of insertion of M. extensor indicis longus. M. flexor meta-
carpi brevis is absent in Coua, Carpo.coccyx, Geococcyx, and Morococcyx (as
well as in Tauraco); it is present in all other cuckoos. examined. It should be
noted that this muscle is present in all parasitic cuckoos and in all of those that
lack pars iliofemoralis of the piriformis muscle. It is present also in Croto-
phaga, Guira, and in Ceuthmochares (in which genus pars iliofemoralis is
vestigial); it is present but very much reduced in size in Phaenicophaeus.
Only in Geococcyx, Morococcyx, and Coua is M. coracobrachialis anterior a
very large muscle, extending upward around the anterior edge of the humerus
and to the dorsal surface of that bone. In the other genera, M. coracobrachi-
alis anterior is a small muscle on the ventral surface of the humerus where it is
closely applied to the capsule of the shoulder joint.
M. pectoralis propatagialis longus et brevis may be entirely tendinous
(Morococcyx, Dromococcyx, Tapera, Guira, Ceuthmochares, Centropus,
Cuculus, Piaya), or it may arise from M. pectoralis as a fleshy bundle (Geo-
coccyx, Crotophaga, Coccyzus, Saurothera, Coua, Carpococcyx, Clamator,
Chrysococcyx, Surniculus, Phaenicophaeus ).
M. deltoideus minor apparently is absent only in Coccyzus and Centropus.
Three patterns in the relative development of Min. pronator superficialis
and pronator profundus are seen in the cuckoos: (1) the two muscles are
about equal in length and both are relatively short Coccyzus, Dromococcyx,
Chrysococcyx, Surniculus; (2) they are about the same length and they extend
nearly to the distal end of the radiusMorococcyx, Guira, Centropus, Ceuth-
Mach 1960
78 THE WILSON BULLETIN VoL 7::, No. I
mochares, Phaenicophaeus; (3) the profundus is longer than the superficialis
Coua, Geococcyx, Crotophaga, Tapera, Paya, Saurothera, Carpococcyx,
Clamator, Cuculus.
Cuculus and Clrysococcyx appear to differ from all other genera in that
they possess an accessory flexor muscle of the pollex, which arises from the
tendon of insertion of M. tensor patagii longus (Berger, 1955b: Fig. 70). In
these two genera and in Surniculus and Phaenicophaeus, M. flexor digitorum
profundus arises on both sides of the insertion of M. brachJails, rather than
only posterior to it as in the other cuckoos.
The relative development of M. flexor carpi ulnaris brevis appears to be
correlated with locomotor habits. The muscle is very short in Geococcyx,
Morococcyx, Dromococcyx, Tapera, Crotophaga, Coua, Carp.ococcyx, Cen-
tropus, Phaencophaeus, and Ceuthmochares. It is relatively long in Coccyzus,
Piaya, Saurothera, Guira, Clamator, ,Cuculus, Chrysococcyx, and Surniculus.
Throughout the discussion above, I have used qualifying terms (e.g., pri-
marily, in general, more arboreal) when referring to cuckoos as being arboreal
or terrestrial. A review of the general habits and the myology of the couas
may be instructive in this regard. The genus Coua contains a well-defined
group of cuckoos that is restricted to Madagascar. Several years ago I (1953a:
53-54) quoted from a letter that Dr. A. L. Rand had written to me about this
genus. He wrote, in part: "The three arboreal species caerulea, cristata, and
verreauxi are geographical replacements. The terrestrial rain-forest species,
reynaudi and serriana, have different food, one insects, one fallen fruit. In
the dry forests and brush are four species, one rather wide-spread (ru[iceps),
but the other three are more restricted. C. coquerel and cursor are. geographi-
cal representatives, and the much larger C. gigas co-exists with C. coquerel
but overlaps the range of C. cursor.
"The group is, to my mind, a terrestrial type. The terrestrial species are at
home, agile and graceful on the ground, swift of foot, and may prefer to
escape danger on foot rather than by flying or seeking concealment. The
arboreal species, and this includes C. caerulea, have carried over as much of
this as possible into their arboreal life.
"The various modifications in range, habitat, and habits seem to be the
result of an intra-group pressure and competition with little competition, in
island isolation, from other groups. They have modified these without modifi-
cation in structure. Thus, to use C. caerulea as an example of an arboreal
cuckoo is a mistake. It is a terrestrial type that has taken to living in trees!
Hence, its resemblance. in leg proportions to Geococcyx is what I would
expect."
Dr. Reynaud Paullan wrote (letter, January 16, 1956) that "Coua ru[iceps
walks along branches, just as the African Musophaga do, moving one leg
Andrew J.
Berger
ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 79
after the other in alternating movements. Only when excited will it run by a
series of sharp jumps on the ground. The Coua walks along the ground if not
afraid, but it will take to the jumping run if at all disturbed."
In view of these statements made by two men who have observed the couas
in the field, it is interesting to report that I found only one difference (not
correlated to size difference) in the royology of five species of Coua, and there
is a possibility of individual variation in this feature. In C. serriana and C.
reynaudii, M. extensor pollicis brevis arises by two separate heads: one from
the tendon of insertion of M. extensor pollicis longus; the other, from the base
of the carpometacarpus adjacent to the insertion of the tendon of M. flexor
carpi ulnaris brevis. M. extensor pollicis brevis has a single head in caerulea,
cristata, and ru/iceps. The remaining wing muscles and the leg muscles
exhibit an identical pattern of development in the five species. We can, indeed,
agree with Dr. Rand that the couas have modified their habits without
extensive modification in royology. This suggests that myological data can be
used with some degree of confidence for determining closeness of relationship,
providing, of course, that other factors are taken into consideration. One does
not find, however, a similar uniformity in the relative lengths of the bones of
the appendages within the genus Coua. Milne-Edwards and Grandidier (1879:
170) commented on differences in development of the leg in various species
of this genus. The following is a free translation of a portion of their com-
ments: One does not perceive any relation between the habits of the birds and
the dimensions of the tarsus; thus, the tarsi are long in Coua caerulea, which
is arboreal, and in C. gigas and C. ru]iceps, which are terrestrial; they are, on
the contrary, short in C. reynaudii and in C. cursor, whose habits are very
different, the first being a climber, and the second a runner. The chief
difference in the leg muscles of the couas is in the increased length of the
tendons in the longdegged forms.
Another example of difference in habit without elaborate modification in
myology may be provided by the genera Geococcyx and Morococcyx. These
two presumably closely related genera exhibit essentially an identical pattern
in the royology of both limbs. The well-known Roadrunner (Geococcyx) is a
terrestrial bird that has a running speed estimated at 10 to 15 mph. Moro-
coccyx also is a "Ground-Cuckoo," but little seems to be known about its life
history or habits. However, Dr. Ernest P. Edwards wrote (letter, January 23,
1953) that Morococcyx "walks along branches somewhat as a dove would
walk along them." Perhaps Morococcyx, also, is a ground-cuckoo that has
taken to living in trees.
A number of other cuckoos are intermediate in locomotor habits. For
example, Delacour (1947:120) characterized the species of Phaenicophaeus
March 1960
80 THE WILSON BULLETIN Vo. 72, No. 1
(in which genus he included all the species of nine of Peters' genera) as being
poor fliers, living among vines and bushes.
The correct interpretation of differences in royology within the family
Cuculidae is complicated because some genera are parasitic and others are not.
If one attempts to use myological formulae as well as breeding behavior in
determining relationships among cuckoos, one is confronted with some puz-
zling situations, even if one excludes "E" from the muscle formulae. Among
New World cuckoos, for example, Tapera is known to be parasitic, and the
meager information available indicates that Dromococcyx, also, is parasitic
(Giai, 1949; Neunteufel, 1951). Tapera, like all of the Old World parasitic
cuckoos except Eudynamis and Scythrops, lacks pars iliofemoralis of the
piriformis muscle, whereas Dromococcyx, Eudynamis, and Scythrops are,
apparently, the only parasitic cuckoos that have what is traditionally consid-
ered the full cuculine muscle formula of ABXYAm. Hence, it would appear
that one must discount either myological data or breeding behavior in deciding
the relationships among the cuckoos. It must be emphasized, however, that we
are handicapped seriously because so little is known about the life histories of
the Neotropical (as well as many other) cuckoos.
Friedmann (1933) suggested that the parasitic habit is "of great antiquity"
in the cuckoos and that Tapera developed its parasitism independently of the
Old World cuckoos. I would suggest that the parasitic habit did not develop
independently in Tapera and Dromococcyx but that "the ancestral Cuculine
stock that reached the Americas brought with it a tendency towards parasitism
(if not an actually developed parasitic habit) which has fructified in Tapera"
(Friedmann, 1933:533) and in Dromococcyx. There are parasitic cuckoos in
both hemispheres that possess the two basic types of cuculine muscle formulae:
(1) AXYAm: Old World--Cuculus, Chrysococcyx, Clamator, Surniculus, etc.;
New World--Tapera. (2) ABXYAm: Old World-Eudynamis, Scythrops;
New World--Dro.mococcyx. Thus, if we are to place any value on morpho-
logical characters, we must assume either that parasitism has developed
independently as many as four times in this one family (which seems highly
unlikely) or that the parasitic habit (or tendency for it) developed in the
primitive cuckoos (all ABXYAm) while still in the Old World ancestral home
of the family.
COMPARATIVE OSTEOLOGY
Moreau recently called attention to the considerable difference of opinion
expressed by the several classifications of the touracos that have been pro-
posed during the past 34 years. I am in no sense prepared to comment on the
classification of the touracos, but I can add some pertinent remarks on the
osteology of the group. Several years ago, I examined the skeletons of 14
touracos, representing four of Moreau's (1958:75) five genera and eight of his
Andrew $. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 81
Berger
18 species: Corythae'ola cristata; Crini[er piscator; Corythaixoides concolor
and personata; Tauraco corythaix (including persa and persa bu[]oni), mac-
rorhynchus, and leucotis donaldsoni.
Verheyen (1956a, 1956b) analyzed the skeletons of touracos and cuckoos in
some detail, and from this and other information taken from the literature
proposed systems of classification for both the touracos and the cuckoos. With
regard to the touracos, Moreau (1958:77) commented: "I understand from
Verheyen (in litt.) that his sub-families are not based on osteological charac-
ters except for the Corytheolinae (which have 20 pre-synsacrals and five
dorso-sacra, compared with nineteen and four respectively in all the other
Musophagidae). The separation of Corytheola in this way might perhaps be
justifiable, but the allocation of the remaining Musophagidae into three sub-
families can hardly be accepted."
The separation of Corythaeola from the other musophagids on the basis of
the number of cervical or presynsacral vertebrae is not justifiable. I studied
two skeletons of Corythaeola cristata: one had 14 cervical and 5 dorsal verte-
brae, or 19 presynsacral vertebrae as in the other musophagids; the other
specimen had 15 cervical and 5 dorsal vertebrae, or 20 presynsacrals, as in the
single specimen examined by Verheyen. This is surely another example of
individual variation in the number of cervical vertebrae within a species; in
my opinion, it has no taxonomic significance. I had for study only one
specimen of Corythaixoides (= Gymnoschizorhis) personata, but it, too, had
20 presynsacral vertebrae. Each of two specimens of Tauraco corythaix persa
(see Moreau, 1958) had 14 cervical vertebrae, but one had 5 dorsal vertebrae,
the other only 4. Hence, one specimen had the typical musophagid number of
19 presynsacral vertebrae, whereas the other had only 18. Despite these few
examples of individual variation, it seems likely that the typical number of
presynsacral vertebrae in the musophagids is 19: 14 cervical and 5 dorsal
vertebrae.
Among the specimens studied by Verheyen, the number of free caudal
vertebrae varied from 6 to 8; among my specimens, from 5 to 7.
The number of ribs that articulate directly with the sternum varies, among
the several genera and within a species, from 3 to 5, with 4 being the usual
number. In one specimen of Corythaixoides concolor, five ribs articulated
with the sternum on the right side, but only four did so on the left.
If we now turn our attention to the Cuculidae, we find that the typical
number of presynsacral vertebrae in most genera is 18: 14 cervical and 4
dorsal vertebrae. However, I found that each of 30 specimens of the genus
Coccyzus had only 13 cervical and 4 dorsal vertebrae, or a total of 17 pre-
synsacral vertebrae; one additional specimen had 14 cervical and 4 dorsal
vertebrae. According to Shufeldt, Clamator glandarius has only 13 cervical
March 1960
82 THE WILSON BULLETIN vol 72, No. 1
vertebrae; Verheyen (1956a:17) indicated that he examined four specimens
of C. levaillantii and found 13 cervical vertebrae in this species; I found 13
cervical vertebrae in my single specimen of C. ]acobinus. Apparently there
are no reliable published data to indicate that any other cuculine genus has 13
cervical vertebrae. Fiirbringer (1888:779), for example, is in error when he
lists 13 cervical vertebrae for Crotophaga, except, of course, as a matter of
individual variation; the genus typically has 14 cervical vertebrae. However, I
found only 13 cervical and 4 dorsal vertebrae in single specimens of Piaya and
Saurothera. One specimen of Guira had 14 cervical and 4 dorsal vertebrae,
whereas a second specimen had 13 cervical and 5 dorsal vertebrae, certainly
an abnormal number for cuckoos. The number of cervical and dorsal verte-
brae within a genus appears to be relatively constant in the cuckoos (Berger,
1956:437), and it seems likely that the usual pattern in Guira is 14 cervical
and 4 dorsal vertebrae. The osteology of only one genus (Coccyzus) can be
considered to have. been studied in sufficient numbers to justify much confi-
dence in numerical data, however, so that much additional information is
needed on the osteology of the cuckoos before the full significance of the data
can be appreciated. Nevertheless, it is interesting to note that, apparently,
Coccyzus, Piaya, Saurothera, and Clamator, each have 13 cervical vertebrae,
and each genus has the same muscle formula of AEXYAm.
In my single specimen of Carpococcyx (United States National Museum No.
223,970), there were 15 cervical and only 3 dorsal vertebrae. There was,
however, a well-developed thoracic rib (which, by definition, articulates dor-
sally with the synsacrum and not with a dorsal vertebra), and this rib had a
direct articulation with the sternum, so that 4 ribs (3 true ribs and 1 thoracic
rib) articulated with the sternum bilaterally. This represents a very unusual
condition, at least among cuckoos. I suspect that it is an anomaly in this one
specimen, and that Carpococcyx, like most other cuckoos, typically has 14
cervical and 4 dorsal vertebrae.
We may summarize some of the pertinent differences in the osteology of the
two groups of birds. Touracos typically have 19 presynsacral vertebrae; most
cuckoos have 18 presynsacral vertebrae, but a few have only 17. Touracos
have a notched atlas; cuckoos, a perforated atlas. Touracos have a single
bony canal in the hypotarsus; cuckoos have two bony canals. The sternum is
double-notched in the touracos; the sternum is either single-notched or
double-notched in cuckoos, sometimes (in older individuals?) it is merely
fenestrate. In the touracos (all?) the clavicles are not fused ventrally to form
a furcula, but the inferior ends are united by a ligament; a furcula is present
in the cuckoos. As pointed out by Pycraft (1903:279) and Verheyen (1956a),
the two dorsal processes of the coracoid are fused to form a bony canal in the
touracos; the two processes do not fuse in the cuckoos. In the touracos, the
Andrew J.
Berger
ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 83
tendon of M. supracoracoideus passes through the canal in the coracoid and
then through the foramen triossenna, which is bounded only by coracoid and
scapula. Lowe (1943) described and illustrated differences in the structure of
the ectethmoid plate and the lacrimal bone between Tauraco and Cuculus.
Some cuckoos (e.g., Centropus) are somewhat intermediate between Cuculus
and Tauraco in the development of these characters, but the over-all pattern
probably is distinctive for the two families (Fig. 2).
CRINIFER
Fig. 2. Lateral view of part of skull of Criniter piscator to emphasize the lacrimal hone.
COMPARATIVE PTERYLOSIS
Forbes (1885:212) remarked that the pterylosis of the Musophagidae "is
peculiar, and gives no aid in determining their affinities." Nitzsch (1867:
PI. 6) illustrated the dorsal and ventral feather tracts of Tauraco erythrolophus
("Musophaga paulina"), and Lowe (1943) illustrated these tracts for Muso-
phaga violacea. I found the feather tracts of Tauraco leucotis donaldsoul to be
essentially the same as those shown in Lowe's figures.
Beddard (1885 and later) placed considerable emphasis on the value of the
ventral feather tracts in determining relationships among the cuckoos, but the
84 THE WILSON BULLETIN March 1960
Vol. 72, No. 1
entire matter is in need of a thorough re-investigation. I have pointed out
elsewhere (1953a, 1955b) that two of Beddard's assumptions were wrong. It
has been demonstrated, however, that there are several patterns of feather
tracts among the Cuculidae. In one group (Cuculus, Cacomantis, Surniculus,
Ceuthmochares, Clamator, Piaya, Coccyzus, Saurothera) there is a single
ventral abdominal tract on each side. In a second group (Chrysococcyx,
Scythrops, Carpococcyx, Centropus, Eudynamis, Phaenicophaeus, Geococcyx,
Crotophaga, Guira, Morococcyx), the ventral abdominal tracts are paired
bilaterally. Furthermore, the ventral cervical tract may bifurcate in the region
of the chin (Piaya, Saurothera, Coccyzus, Phaenicophaeus), about midway
down the neck ( Cuculus, Clamator, Tapera, Crotophaga, Guira, Carpococcyx,
Geococcyx, Morococcyx), or at the base of the neck (Centropus, Ceuthmo-
chares, and probably Surniculus). Beddard (1898b:280) separated certain
Old World cuckoos from certain New World cuckoos because the ventral
cervical tract in the New World cuckoos is "double at [its] commencement."
This latter group is the one for which I have stated that the ventral cervical
tract bifurcates in the region of the chin. There is some minor variation in
these tracts among the four genera that ! have listed above, and this character
is perhaps most striking in the Old World genus Phaenicophaeus. In Phaeni-
cophaeus there are two interramal feather tracts, separated by a median
apterium, and each tract is bounded laterally by a pronounced marginal
apterium.
Similarly, there are differences in the pattern of the dorsal feather tracts.
In Cuculus canorus, for example, the dorsal cervical tract is continuous with
the interscapular tract, and there is a large median apterium in the latter tract
(Lowe, 1943: Fig. 6). In Coua, Geococcyx, Morococcyx, Carpococcyx, Guira,
and Centropus, however, the dorsal cervical tract ends abruptly at the base of
the neck, and the interscapular tract begins as two independent, anteriorly-
directed tracts, which fuse posteriorly to form a single median pelvic tract.
There are some differences among these genera and it seems pertinent to
describe both the dorsal and ventral tracts of Carpococcyx. I have already
described these tracts in Coua caerulea and C. ruficeps (1953b, 1954); Shu-
feldt (1886) illustrated the pterylosis of Geococcyx; and Beddard (1885)
illustrated feather tracts of Centropus, Eudynamis, Piaya, and Cacomantis.
Shelford (1900) described and illustrated the pterylosis of the embryo and
nestling of Centropus sinensis, and Hartley (in Beebe et al., 1917) illustrated
the feather tracts in both the embryo and adult of Crotophaga ani.
A median frontal apterlure is wanting and the feathers of the capital feather
tract are a little more sparse in Carpococcyx radiceus than they are in Coua.
A large lateral apterlure surrounds the orbit. The spinal cervical tract ends
abruptly at the junction of the neck and thorax, and there is a relatively wide
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE
Berger
85
,0 '
J ß o
J J
FIc. 3. Dorsal feather tracts of Carpococcyx radiceus.
March 1960
86 THE WILSON BULLETIN Vol. 72, No. 1
(20 mm.) apterlum between this tract and the interscapular tracts. These
tracts (Fig. 3) differ from those found in other cuckoos thus far studied, as
pointed out by Beddard (1901:201) but not precisely in the way that he
stated. There does appear to be a short, narrow, V-shaped median apterium
between the two interscapular tracts anteriorly, although in shape and extent it
is unlike that in Coua (Fig. 4). The two interscapular tracts are continu-
ous posteriorly with the dorsal spinal tracts, which join to form a single,
median pelvic tract. In addition, lateral dorsal tracts, essentially single-rowed,
pass caudally from the anterolateral edge of the dorsal spinal tracts to the area
over the posterior illac crests. There is an almost featherless area between
these bilateral tracts and the median pelvic tract.
Turning now to the ventral feather tracts of Carpococcyx, we note first that
the marginal apteria are very small, if present at all (an incision had been
made along the mandible, thus making this determination somewhat unre-
liable). The presence or absence of the marginal apteria on either side of the
interramal feather tract was one of the diagnostic features used by Beddard
(1898b:281): (1) ventral tract occupying the entire space between the
mandibles; or (2) ventral tract occupying only the median region between
the mandibles, that is, the interramal tract is bounded by marginal apteria.
Judging from what I have seen to date, I would say that Beddard's criteria for
separating the Old World Centropus (= "Pyrrhocentor") and Coua from the
New World Geococcyx, Crotophaga, and Guira are untenable. The pattern in
Cuculus certainly is not precisely like that illustrated by Lowe (1943) in that
there are definite small apteria on either side of the interramal tract; nor, in
this character, is there such a close similarity between Cuculus and Musophaga
as Lowe illustrated. The marginal apteria in Centropus are like those in Cotta
caerulea and C. rujiceps. It seems to me that we are dealing here with the
degree to which the marginal apteria are developed. The narrowest interramal
tract, and therefore the widest marginal apteria, seems to be found in Croto-
phaga. Thus between Crotophaga and Cuculus or Centropus there is a marked
difference in the development of the marginal apteria, but Tapera (Figs. 5 and
6), Guira, Geococcyx, and Morococcyx are intermediate between the two ex-
tremes, And, as pointed out above, Phaenicophaeus and certain other genera
have two interramal tracts, each of which is bounded laterally by a marginal
apterium.
The ventral cervical tract of Carpococcyx bifurcates at about the midlength
of the neck. Opposite the head of the furcula, two rows of feathers pass
dorsocaudad to join the hmneral feather tract. The ventral tract bifurcates at
about the level of the anterior end of the sternum, forming an inner and an
outer ventral abdominal tract bilaterally. The inner tract is composed of a
double row of feathers anteriorly, but of a single row in the posterior half of
Andrew J.
Berger
ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE
87
Fig. 4. Dorsal feather tracts of nestling Coua ru[iceps.
88
THE WILSON BULLETIN
March 1960
Vol. 72, No. 1
Fig. 5. Dorsal feather tracts of Tapera naevia.
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE
Berger
89
Fic. 6. Ventral feather tracts of Tapera naevia.
March 1960
90 THE WILSON BULLETIN Vol. 72, No.
its extent; this tract terminates a short distance (about 6 min.) in front, and a
little anterolateral, of the vent. The outer abdominal tract anteriorly is com-
posed of three rows of four feathers each, but throughout most of its extent it
is composed of a single row of feathers. This row ends about midway between
the posterior margin of the sternum and the vent; it does not turn inward to
meet the inner abdominal tract. From the lateral margin of the anterior part
of the outer abdominal tract, a single row of feathers passes dorsolaterad to
the posteoventral surface of the arm (humerus).
There are 10 primaries in Carpococcyx. The wing is eutaxic. There are
apparently 11 secondaries; the outer eight have large quills, whereas the
inner three are progressively smaller and are more like coverts in size, but they
are in direct line with the outer secondaries. Beddard considered only the
distal eight feathers to be secondaries. There are five alula quills; the molt
was in progress when the specimen was collected so that it was not possible to
determine the relative lengths of the alula feathers. A relatively well-developed
(40 ram. long) carpal remex was found bilaterally, but no evidence of the
carpal covert. There are 10 rectrices and 10 upper tail coverts. The femoral
tract laterally is developed as in Coua ru/iceps; medially, the thigh is almost
devoid of feathers. The crural tract is best developed anteriorly, where there is
a triple row of feathers proximally (somewhat irregular in pattern) and a
single row distally; there are three rows of widely-spaced feathers on the
lateral side of the crus; medially, there are a few scattered feathers; and
posteriorly, there is a single row of small feathers. The crural tracts do not
extend downward on to the tarsometatarsus, as they do in Cuculus, Chryso-
coccyx, and Clamator. Sharpe (1873:579) spoke of Cuculus, Chrysococcyx,
and Clamator as having "accipitrine thigh-feathers," and he reported that the
rim around the external naris is swollen in these genera; Surniculus lugubris
also exhibits both of these characters.
Apparently all cuckoos have 10 primaries. Most cuckoos have either 9 or 10
secondaries; Scythrops is said to have 13; Carpococcyx and Centropus appear
to have 11; and the nestling Coua ru]iceps seems to have 12 (see, however,
the discussion by Berger and Lunk, 1954). Guira and Cro.tophaga have 8
rectrices; all other cuckoos are said to have 10.
Despite the variation in pattern of both the dorsal and the ventral feather
tracts among the cuckoos, no cuckoo yet studied has feather tracts (particu-
larly dorsally) like those of the touracos. The oil gland is tufted in touracos;
it is nude in cuckoos. An aftershaft is present in touracos; it is absent or
vestigial in cuckoos (Miller, 1924). Well-developed "eyelashes" are charac-
teristic of cuckoos. The eyelashes may be featherlike, as in Cuculus, Chryso-
coccyx, Clamator, and Surniculus, or they may be "hairlike," as in Coua,
Geococcyx, Morococcyx, Dromococcyx, Tapera, Guira, Crotophaga, Piaya,
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 91
Berger
Saurothera, Ceuthmochares, Centropus, and Phaenicophaeus. The small, black,
hairlike eyelashes are least conspicuous in Phaenicophaeus pyrrhocephalus;
those of Dasylophus (= Phaenicophaeus?) superciliosus are said to be large
and scarlet. Eyelashes are absent in Tauraco leucotis donaldsoni, and both
eyelids are covered by fleshy caruncles which are especially prominent dorsal
to the upper eyelid. There appear to be seven alula quills attached to the
pollex in Tauraco; these increase in length from proximal to distal: No. 1 is
23 mm. long; No. 2, 28 ram.; No. 3, 33 mm.; No. 4, 36 mm.; No. 5, 44 mm.;
No. 6, 50 mm; No. 7, 52 mm. Cuckoos apparently have either four or five
alula quills. The innermost, the outermost, or one of the intermediate feathers
may be the longest in the series.
THE CUCULINE SYRINX
There are few anatomical structures throughout the families of birds that
need study as badly as the syrinx. Except for a very few isolated instances,
our knowledge of both the passerine and the nonpasserine syrinx is based
largely on work published during the last century (e.g., Mtiller, 1878). Such
pioneers as Mtiller, Garrod, Forbes, Ftirbringer, and Gadow were remarkably
astute scholars of the old school and any thorough anatomical study must
begin with a careful analysis of their papers. Nevertheless, the work of the
past needs to be corroborated and extended, using both gross and microscopic
techniques. We need detailed comparative studies of the syrinx interpreted in
terms of modern concepts of anatomy and of systematics. For such studies,
one needs a number of specimens of each species in order to work out all
relationships carefully. The intrinsic syringeal muscles usually are very small
and they often are poorly preserved in alcoholic specimens. In addition to
gross descriptions, microscopic analyses and embryological studies of develop-
mental stages of the syrinx in both sexes would undoubtedly add much to our
understanding of the significance of the adult configuration of the syrinx and
its relation to muscles and to the tracheobronchial tree.
I can do little more here than verify statements in the literature and sum-
marize present knowledge of the cuculine syrinx. The following cuckoos are
known to have a bronchial syrinx: Centropus, Carpococcyx, Coua, Geococcyx,
Morococcyx, Dromococcyx, Crotophaga, and Guira. Within this group, how-
ever, there is a considerable difference as to the location of the syrinx within
the bronchi, as pointed out by Beddard (1885:170-174). Beddard reported
(1901:202) that "the syrinx of Carpococcyx . . . is the most purely bronchial
syrinx that exists among the Cuculidae; it is even more exaggerated than that
of Crotophaga," and he added that the syringes of Centropus and Geococcyx
"present us with an intermediate stage between the typical tracheo-bronchial
syrinx of the Phoenicophainae and the extraordinarily specialized bronchial
March 1960
92 THE WILSON BULLETIN Voi. 72 No. x
syrinx of Crotop,haga and of Carpococcyx." The syrinx of Centropus super-
ciliosus (Fig. 11), however, approaches the "exaggerated" condition found in
Carpococcyx (Beddard, 1901:203), and it appears to be different from the
syrinx of Centropus ateralbus (Beddard, 1885:172). This suggests that a
comparative study of the syrinx among the many species of Centropus would
be a fruitful project.
All other cuckoos studied, including Tapera, have a tracheobronchial syrinx.
Beddard remarked (1902:605-606) that the syrinx of Rhamp,hococcyx
(= Phaenicophaeus according to some writers) exhibits "a hint of a develop-
ment into the bronchial syrinx of the Centropine .Cuckoos." Hence, as with
other anatomical information, we seem to know just enough about the anatomy
of the cuculine syrinx to suggest many intriguing problems for further study
and not enough to understand completely the significance of the data that are
available. My material was not suitable for microscopic study, and I present
here figures (Figs. 7-11) illustrating the gross differences in the syrinx of
one series of cuckoos. A knowledge of the finer structure of the cuculine
syrinx would undoubtedly be instructive. It should be noted that M. sterno-
trachealis inserts very near the tracheal bifurcation in Centropus, whereas it
inserts some distance from the bifurcation in the other genera. Only in the
syrinx of Crotop,haga (a female) did I find two pairs of intrinsic syringeal
muscles; there is a single pair of intrinsic muscles in the other genera.
Discussion
There has been a tendency by some contemporary writers to state flatly that
certain anatomical features in birds are so stable in an evolutionary sense that
similarities in these characters "prove" relationship among the birds con-
cerned. One may make such a basic assumption about a single character and
then proceed to "prove" all sorts of inter-relationships that then permit the
construction of an elaborate phylogenetic tree, which the author may state or
imply is to be accepted without qualification or question. Most often such
authors base their conclusions on a smattering of information about a particu-
lar anatomical character. When, for example, an author purports to "explain"
the course of evolution in all oseine passeriform birds on the basis of a single
bone of the skull and on the size of the birds, I believe that we have reached
the ultimate in this type of systematic approach, one which might aptly be
termed "the pseudo-anatomical approach." Other examples could be cited.
The naive confidence in, and the exaggerated positive interpretation of, inade-
quate data expressed by such authors does a great disservice to anatomical
studies and to the application of anatomical information to avian taxonomy.
I doubt that we know enough about avian anatomy or about the inter-
relationships between structure and function--and all of the related problems
^narew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE
Berger
93
FIe. 7
FI(;. 8
Fro. 9
Fc. 10
FG. 11
Tracheobronchial syrinx of Piaya cayana.
Bronchial syrinx of Guira guira.
Bronchial syrinx of Morococcyx erythropygus.
Bronchial syrinx of Crotophaga sulcirostris.
Bronchial syrinx of Centropus superciliosus.
94 THE WILSON BULLETIN March 1960
Vol. 72, No. 1
of convergence and divergence among birds--to justify such confidence.
Miller (1937:58-59), Rinker, and others have discussed this problem and that
of the inequality of values to be assigned to anatomical features in the study of
both birds and mammals. Thus, Rinker (1954:117), in speaking of rodents,
wrote: "Even the segregation of characters into adaptive and nonadaptive
categories, or a classification based on current survival value, is, in my
opinion, a highly subjective approach. I do not believe that the present state
of knowledge will allow this approach to be used with any appreciable degree
of certainty. It appears to me that success in such an attempt would require
much greater insight into the relationships of structure and function, into the
basis of habitat preferences (or restrictions), and into many more phenomena
than is possible at present." It does not seem to me to be an exaggeration to
make a similar statement about birds, particularly when it seems that we know
less about the significance of anatomical characters in birds than in mammals.
Moreover, we can agree wholeheartedly with Enlow and Brown (1958:221)
when they caution: "The projection of preconceived concepts and prejudices
of evolutionary mechanics into the interpretation of observations is a consid-
eration that must consciously, and continually, be recognized and evaluated.
It is a constant temptation to look for, and thereby find, a logical, uni-linear
assemblage of structural changes that conform with and confirm some pre-
existing idea on how things should be happening."
Before summarizing the data presented in this paper, it will be instructive to
consider Peters' (1940) classification of the cuckoos. There seems to be
general agreement that Peters' classification is unsatisfactory. He had little
published information on cuculine anatomy to rely on, and I suspect that his
system was based primarily on two factors: parasitism and geographical
distribution. An analysis of his subfamilies will give further insight into some
of the problems involved and will reveal where further work is needed.
Peters' subfamily Cuculinae is composed of Old World cuckoos that have
parasitic breeding habits. For most of the included genera the muscle formula
is AXYAm, but for Eudynamis and Scythrops it is ABXYAm. Clamator has
13 cervical vertebrae; so far as is known, the other genera have 14. There is a
single ventral abdominal feather tract on each side in Cuculus, Cacomantis,
Clamator, and Surniculus; these tracts are paired bilaterally in Chrysococcyx,
Scythrops, and Eudynamis. Baker (1927) erected the subfamily Eudynaminae
for the single genus Eudynamis.
An analysis of Peters' subfamily Phaenicophaeinaewhich he admitted was
a "catch-all" group--is even more interesting. He included in this subfamily
three nonparasitic American genera and nine nonparasitic Old World genera.
Each of the American genera (Coccyzus, Piaya, Saurothera) has a muscle
formula of AEXYAm, a single ventral abdominal feather tract bilaterally, and,
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 95
Berger
apparently, 13 cervical vertebrae. Six of the nine Old World genera are known
to have the muscle formula of ABXYAm. Of the two genera studied for this
paper, Ceuthmochares has a single ventral abdominal feather tract bilaterally,
whereas Phaenicophaeus has paired tracts bilaterally. Both genera have 14
cervical vertebrae and a muscle formula of ABEXYAm. It should be pointed
out that Delacour and Mayr (1945) did not recognize eight of the nine Old
World genera (including Ceuthmochares) listed by Peters, and, therefore,
placed all of the species in the genus Phaenicophaeus. Bannerman (1933), on
the other hand, placed Ceuthmochares in his subfamily Centropodinae.
The Crotophaginae (Crotophaga and Guira) are New World cuckoos that
have eight rectrices, a muscle formula of ABEXYAm, and social nesting
habits.
Peters' subfamily Neomorphinae contains one Old World genus (Carpo-
coccyx) and five New World genera. All except Tapera (AXYAm) have a
muscle formula of ABXYAm. Tapera and Dromococcyx are parasitic; so far
as is known, the other genera are not parasitic, although Makatsch (1955:34)
suggested that Neomorphus might have parasitic nesting habits (but see, Sick,
1949). Shelley (1891:423) placed Tapera (="Diplopterus") and Dromo-
coccyx in the subfamily Diplopterinae (= Taperinae), whereas Gadow and
Selenka (1893:214) put Tapera in their subfamily Cuculinae (parasitic,
AXYAm, syrinx tracheobronchial).
The subfamily Couinae presumably was established because the single genus
is restricted to Madagascar. There is little in the internal anatomy to separate
Cotta from Geococcyx or Morococcyx.
Peters' last subfamily, the Centropodinae, contains the single Old World
genus Centropus with some 27 species and 55 subspecies. These are wide-
spread (from Australia, the Philippines, and China to Africa and Madagas-
car), nonparasitic cuckoos that have a very long, straight and sharp claw on
the hallux (of many, but not all, of the species), thus the name of Lark-heeled
Cuckoos. The coucals build a bulky globular nest that has a side entrance
(Delacour, 1947; Loke, 1953), but I suspect that most authors have erected
the subfamily primarily on the basis of the long claw on the hallux.
Both Beddard (1898b) and Pycraft (1903) adopted several anatomical
features--leg-muscle formulae, patterns of the feather tracts, skeletal charac-
ters, and the structure of the syrinx--in an attempt to determine cuculine
relationships, but, as in the present paper, each author had information from
less than half of all the genera. Beddard placed primary emphasis on the
syrinx and the leg-muscle formulae (and omitted skeletal characters) in
assigning cuckoos to three subfamilies. Pycraft omitted consideration of
feather tracts and used skeletal characters (sternum and pectineal process)
March 1960
96 TIlE WILSON BULLETIN Vol. 72, No. 1
and the syrinx in arranging the genera. The divergence in the placement of
genera between these two systems was considerable.
The difficulty of deciding relationships among the genera of cuckoos seems
to result largely from the fact that anatomical characters often presumed to be
very stable in an evolutionary sense exhibit several patterns among the
cuckoos. Thus, the muscle formula may be AXYAm or ABXYAm; the syrinx
may be bronchial or tracheobronchial; the sternum may be single-notched or
double-notched; there may or may not be an apterium between the dorsal
cervical and the interscapular feather tracts; and the ventral abdominal
feather tracts may be either single or double bilaterally. Similarly, parasitic
breeding habits have developed in genera possessing both types of muscle
formula, both types of syrinx, both types of sternum, both types of ventral
feather tracts, and, perhaps, both types of dorsal feather pattern. Moreover,
genera with a bronchial or a tracheobronchial syrinx, those with a single-
notched or a double-notched sternum, etc., are found in both the Old World
and the New World. These points may be illustrated by constructing a key
using such anatomical features rather than using only parasitic breeding
habits, geographical distribution, or the appearance of the study skin as a
basis for classification. It is obvious that an author can focus the reader's
attention on certain relationships by the relative weight placed on anatomical
characters in constructing such keys. To demonstrate this point, I present
three different keys (Tables 1-3), which embody anatomical features used by
Beddard and Pycraft, as well as other features discussed in this paper; by
using information reported by Becldard, I have added a few genera not
examined for this paper. It is obvious, also, that the data in the tables can be
arranged in several other ways.
If we consider only Table ! for the moment, we can say, in addition, that
the genera in Group I have a well-developed peroneus longus muscle; the
genera in Group II, except for Tapera, have a weakly-developed peroneus
longus muscle. M. iliacus ("E") is absent in the Old World genera of Group I.
All the genera in Group I have a bronchial syrinx, bilaterally paired ventral
abdominal feather tracts, and 14 cervical vertebrae; most of the genera are
nonparasitic, but two have developed social nesting habits. Sick (1953a,
1953b) reported that Tapera and Dromococcyx are similar in certain external
features and in general habits, but the two genera are unlike in internal char-
acters. One would conclude that the superficial similarities in appearance are
the result of convergence.
I mentioned above (p. 90) that four genera (Cuculus, Chrysococcyx,
Clamator, and Surniculus) have "accipitrine" thigh feathers and raised narial
rims, but these genera differ considerably in other morphological features. All
four genera have parasitic nesting habits.
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 97
Berger
After examining the tables, one can see why it would be much easier to
classify the cuckoos on the basis of breeding habits or on geographical dis-
tribution rather than on the basis of internal anatomical characters.
Still other anatomical features of the cuckoos need to be studied: the
morphology of the head, particularly the jaw muscles and nerves (see, for
example, Barnikol, 1953, and Starck, 1959); a comparative study of the
viscera, the vascular system, and the brachial and lumbosacral plexuses.
98 THE WILSON BULLETIN March 1960
Vol. 72, No. 1
TABLE 1
ANATOMICAL RELATIONS OF CUCKOOS
Group I. Bronchial syrinx; ABE(_+)XYAm; ventral abdominal feather tracts double
bilaterally; 14 cervical vertebrae.
1. Sternum double-notched; dorsal cervical and interscapular feather tracts
separated by an apterium; M. flexor hallucis longus inserts on hallux.
a.M. flexor metacarpi brevis absent: Coua (E absent), Geococcyx.
Morococcyx, Neomorphus ?
b. M. flexor metacarpi brevis present;
parasitic; E?; dorsal feather tracts?: Dromococcyx
2. Sternum single-notched
a.M. flexor hallucis longus inserts on ha]lux
1) dorsal cervical tract is continuous with
interscapular tract; social nesting habits: Crotophaga
2) dorsal cervical tract separated from
interscapular tract by an apterium
a) M. flexor metacarpi brevis present;
social nesting habits: Guira
b) M. flexor metacarpi brevis absent;
E absent; nonparasitic: Carpococcyx
b. M. flexor hallucis longus does not insert on
hal]ux; E absent; nonparasitic: Centropus
Group II. Tracheobronchial syrinx; dorsal cervical feather tract is continuous with
interscapular tract.
1. ABE (__+ ?) XYAm
a. Sternum double-notched; nonparasitic; E present.
1) ventral abdominal tracts double
bilaterally: Phaenicophaeus
2) ventral abdominal tracts single
bilaterally: Ceuthmochares
b. Sternum single-notched; parasitic; ventral
abdominal feather tracts double bilaterally
(from Beddard): Scythrops, Eudynamis
2. AEXYAm; sternum double-notched; ventral abdominal tracts single.
a. 14 cervical vertebrae; parasitic: Tapera, Surniculus
b. 13 cervical vertebrae
1) parasitic: Clamator
2) nonparasitic: Coccyzus, Saurothera, Piaya
3. AXYAm; parasitic
a. Sternum double-notched; ventral abdominal
tracts double: Chrysococcyx
b. Sternum single-notched; ventral abdominal
tracts single: Cucu[us, Pachycoccyx
Andrew
Berger
ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 99
TABLE 2
ANATOMICAL RELATIONS OF CUCKOOS
Group I.
ABXYAm.
1. Bronchial syrinx; ventral abdominal feather tracts double; 14 cervical
vertebrae.
a. Dorsal cervical and interscapular feather tracts
are separated by an apterium.
1) M. flexor hallucis longus does not insert
on hallux; sternum single-notched; M.
flexor metacarpi brevis present; non-
parasitic: Centropus
2) M. flexor hallucis longus inserts on hallux.
a) Sternum double-notched
1--M. flexor metacarpi brevis absent;
nonparasitic:
Coua, Geococcyx,
Morococcyx, Neomorphus ?
2--M. flexor metacarpi brevis present;
parasitic; dorsal feather tracts ?: Dromococcyx
b) Sternum single-notched
IM. flexor metacarpi brevis absent;
nonparasitic: Carpococcyx
2--M. flexor metacarpi brevis present;
social nesting habits: Guira
b. Dorsal cervical tract continuous with inter-
scapular feather tract; sternum single-
notched; social nesting .habits: Crotophaga
2. Tracheobronchial syrinx; dorsal cervical tract is continuous with inter-
scapular feather tract.
a. Sternum single-notched; parasitic; ventral
abdominal feather tracts double: Scythrops, Eudynamis
b. Sternum double-notched; nonparasitic
1) ventral abdominal tracts double: Phaenicophaeus
2) ventral abdominal tracts single: Ceuthmochares
Group II. AXYAm; tracheobronchial syrinx; dorsal cervical tract continuous with
interscapular feather tract.
1. Sternum double-notched
a. Ventral abdominal tracts single bilaterally
1) 14 cervical vertebrae; parasitic: Tapera, Surniculus
2) 13 cervical vertebrae
a) nonparasitic: Coccyzus, Saurothera, Piaya
b) parasitic: Clamator
b. Ventral abdominal tracts double; parasitic: Chrysococcyx
2. Sternum single-notched; ventral abdominal
tracts single; parasitic: Cuculus
lOO
THE WILSON BULLETIN
March 1960
Vol. 72, No. 1
TaBus 3
ANATOMICAL RELATIONS OF CUCKOOS
Group I.
Sternum double-notched.
l. Bronchial syrinx; ABE(+)XYAm
a.M. flexor metacarpi brevis absent; nonpara-
sitic:
Cotfi G G1/2ococcy.
Morococcyx, Neomorphus ?
b. M. flexor metacarpi brevis present; para-
sitic: Dromococcyx
2. Tracheobronchial syrinx
a. ABEXYAm; nonparasitic
1) Ventral abdominal tracts double: Phaenicophaeus
2) Ventral abdominal tracts single: Ceuthmochares
b. AEXYAm; ventral abdominal tracts single
1) 14 cervical vertebrae; parasitic: Tapera, Surniculus
2) 13 cervical vertebrae
a) parasitic:
b) nonparasitic:
c. AXYAm; ventral tracts double; parasitic:
Group II. Sternum single-notched.
1. Bronchial syrinx
a.M. flexor hallucis longus inserts on hallux
1) Dorsal cervical tract continuous with
interscapular tract; social nesting habits: Crotophaga
2) Dorsal cervical tract separated from
interscapular tract by an apterinm.
a) Flexor metacarpi brevis present; so-
cial nesting habits: Guira
b) M. flexor metacarpi brevis absent;
E absent; nonparasitic: Carpococcyx
b. M. flexor hallucis longus does not insert on
hallux; E. absent; nonparasitic: Centropus
2. Tracheobronchial syrinx; parasitic
a. ABXYAm; ventral abdominal tracts double: Scythrops, Eudynamis
b. AXYAm; ventral abdominal tracts single: Cuculus, Pachycoccyx
Clamator
Coccyzus, Saurothera, Piaya
Chrysococcyx
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE
Berger
101
TABLE 4
ANATOMICAL CHARACTERS OF CUCKOOS AND TOURACOS
Cuculidae Tauraco leucotis
1. ABDFGXYAmV
1. AFGXYAm, AEFGXYAm, ABFGXY
Am, ABEFGXYAm
2. Expansor secundariorum is cuculine
3. 2 bony canals in hypotarsus
4. 17 or 18 presynsacral vertebrae; 4 dor-
sal vertebrae typical
5. Perforated atlas
6. Typical furcula present
7. No bony canal formed by coracold
8. Lacrimal bone is "cuculine"
9. Oil gland is nude
10. Aftershaft is absent or vestigial
11. Eyelashes are characteristic
12. Pattern of dorsal feather tracts is dis-
similar
2. Expansor secundariorum is ciconine
3. 1 bony canal in hypotarsus
4. 19 presynsacral vertebrae; 5 dorsal ver-
tebrae typical
5. Notohed atlas
6. Furcula absent
7. Bony canal is formed by the dorsal
processes of the coracold
8. Lacrimal bone is "musophagine"
9. Oil gland is tufted
10. Aftershaft is present
11. Eyelashes are absent
LITERATURE CITED
BAKER, E. C.
1927 The fauna of British India. Birds, Vol. 4. Taylor & Francis, London.
BANNERMAN, D. A.
1933 The birds of tropical West Africa. Vol. 3. London.
BARNIKOL, A.
1953 Vergleichend anatomische und taxonomisch phylogenetische Studien am Kopf
der Opisthocomlformes, Musophagidae, Galli, Columbae und Cuculi. Zool.
Yahrb., 81:487-526.
BEDDARD, F. E.
1885 On the structural characters and classification of the cuckoos. Proc. Zool. $oc.
London, 1885:168-187.
1898a On the anatomy of an Australian cuckoo, $cythrops novae-hollandiae. Ibid.,
1898:4449.
1898b The structure and classification of birds. Longmans, Green & Co., London.
1901 On the anatomy of the Radiated Fruit-Cuckoo (Carpococcyx radiatus). Ibis,
1901:200-214.
1902 On the syrinx and other points in the structure of Hierococcyx and some allied
genera of cuckoos. Ibis, 1902:599-608.
BEEBE, W., G. I. HARTLEY, AND P. G. HOWES
1917 Tropical wild life in British Guiana. New York Zool. Soc., 504 pp.
BERCER, A. J.
1952 The comparative functional morphology of the pelvic appendage in three genera
of Cuculidae. ,4mer. Midl. Nat., 47:513-605.
1953a On the locomotor anatomy of the Blue Coua, Coua caerulea. ,4uk, 70:49-83.
1953b The pterylosis of Coua caerulea. Wilson Bull., 65:12-17.
March 1960
102 THE WILSON BULLETIN vo. 72, No. 1
1954 The myology of the pectoral appendage of three genera of American cuckoos.
Misc. Publ. Univ. Mich. Mus. Zool. No. 85.
1955a Suggestions regarding alcoholic specimens and skeletons of birds. Auk, 72:
300-303.
1955b On the anatomy and relationships of glossy cuckoos of the genera Chrysococcyx,
Lampromorpha, and Chalcites. Proc. U.S. Natl. Mus., 103, No. 3335.
1956 Anatomical variation and avian anatomy. Condor, 58:433-441.
1959 Leg-muscle formulae and systematics. Wilson Bull., 71:93-94.
BERGER, A. J., arid W. A. LUNK
1954 The pterylosis of the nestling Coua ru/iceps. Wilson Bull., 66:119-126.
DELA1/2OUR, J.
1946 Notes on the taxonomy of the birds of Malaysia. Zoologica, 31:1-8.
1947 Birds of Malaysia. Macmillan Co., New York.
DELaCOUR, J., arid E. MAYR
1945 Notes on the taxonomy of the birds of the Philippines. Zoologica, 30:105-117.
ENLOW, D. H., arid S. O. BRow
1958 A comparative histological study of fossil and recent bone tissues. Part IIL
Texas Jour. Sci., 10:187-230.
FISHER, H. I.
1957 The function of M. depressor caudae and M. caudofemoralis in pigeons. Au]c,
74:479486.
FISHER, H. I., AND D.C. GOODMAN
1955 The myology of the Whooping Crane, Grus americana. Illinois Biol. Monog., 24.
FORBES, W. A.
1885 The collected scientific papers of the late William Alexander Forbes. R. H.
Porter, London.
FRIEDMA rN, H.
1933 A contribution to the life-history of the Crespin or Four-winged Cuckoo, Tapera
naevia. Ibls, 1933:532-539.
FiRBRINGER, M.
1888 Untersuchungen zur Morphologic und Systematik der V6gel .... 2 Vols. Van
Holkema, Amsterdam.
GaDow, H., erDE. SELENKA
1891--1893 Aves. In Bronn's Klassen und Ordnungen des Thier-Reichs. 2 Vols.
Leipzig.
GaRROD, A. H.
1881 The collected scientific papers of the late Alfred Henry Garrod. R. H. Porter,
London.
GIaI, A. G.
1949 Sobre un ejemplar de Dromococcyx pavoninus Pelzeln. E1 Hornero, 9:84-87.
HUDSON, G. E.
1937 Studies on the muscles of the pelvic appendage in birds. 4mer. Midl. Nat.,
LOKE, W.
1953
LOWE, P.
1943
18:1-108.
T.
Some breeding birds of Singapore. Jour. Bombay Nat. Hist. Soc., 51:590-596.
R.
Some notes on the anatomical differences obtaining between the Cuculidae and
the Musophagidae .... Ibis, 1943:490-515.
Andrew J. ANATOMY OF CUCULIDAE AND MUSOPHAGIDAE 103
Berger
MAKATSCH, W.
1955 Der Brutparasitismus in der Vogelwelt. Neumann, Radebeul and Berlin, 236 pp.
MAYa, E.
1944 Birds collected during the Whitney South Sea Expedition. 54. Amer. Mus.
Novit. No. 1269.
MAYa, E., AND D. AMADON
1951 A classification of recent birds. Ibid., No. 1496.
MILLER, A. H.
1937 Structural modifications in the Hawaiian Goose (Nesochen sandvicensis) ....
Univ. Cccli/. Publ. ZooL, 42:1-80.
MILLER, W. DEW.
1924 Variations in the structure of the aftershaft and their taxonomic value. Amer.
Mus. Novit. No. 140.
MILNE-EDWARDS, A., AND A. GRANDIDIER
1878-1879 Histoire physique, naturelie et politique de Madagascar. Vols. 12 and 13.
Histoire naturelie des oiseaux. A. Grandidler, Paris.
MITCHELL, P. C.
1913 The peroneal muscles in birds. Proc. Zool. Soc. London, 1913:1039-1072.
MOaEAU, R. E.
1938 A contribution to the biology of the Musophagiformes, the so-called Plantain-
Eaters. Ibis, 1938:639%71.
1958 Some aspects of the Musophagidae. Ibis, 100:67-112; 238-270.
Mi)LLER, J.
1878 On certain variations in the vocal organs of the Passeres that have hitherto
escaped notice. Clarendon Press, Oxford, 74 pp.
NEUNTEUFEL, A.
1951 Observaciones sobre el Dromococcyx pcvoninus Pelzeln y el parasitismo de los
cuculidos. El Hornero, 9:288290.
NITZSCH, C. L.
1867 Pterylography. Ray Society, London.
PETERS, J. L.
1940 Gheck-list of birds of the world. Vol. 4. Harvard Univ. Press, Cambridge.
PYCaAET, W. P.
1903 Contributions to the osteology of birds. Part V[. Cuculiformes. Proc. ZooL
Soc. London, 1903:258 291.
RINS:Ea, G. C.
1954 The comparative royology of the mammalian genera Sigmodon, Oryzomys, Neo.
tome, and Peromyscus (Cricetinae) .... Misc. Publ. Univ. Mich. Mus. Zool.
No. 83.
SAaPE, R. B.
1873 On the Cuculldae of the Ethiopian Region. Proc. Zool. Soc. London, 1873:
578-624.
SHELFORD, R.
1900 On the pterylosis of the embTyos and nestlings of Centropus sincrisis. Ibis,
1900:654-667.
SHUFELDT, R. W.
1886 Contributions to the anatomy of Geococcyx caliy;orrdanus. Proc. Zool. Soc. Lon-
don, 1886:466491.