The Coracoid Sulci Are Peculiar in Their Depth and Narrowness. the Dorsal

Home , Aquilinae, Henicopernis, Herpetotherinae, Neophrontops, Phalcoboenus, Polihierax

M. JOLLIE, FALCONIFORMES(part IV) 1(201)

The coracoid sulci are peculiar in their depth and narrowness. The dorsal margins of the sulci have distinct, anteriorly facing coracoid pads, correlated with the reduced posterior flange of the coracoid. In Henpetothenee there are large, separate, lateral processcs, while Mieraetun, Miluago, Polyborua, and some species of Faleo have these processes noved toward the midline; a single median dorsal spine is found typically in Faleo (fig. 166). These processes are produced by ossifications in the origins of the supraeo"aeoi-deus muscles. Sna1l lateral processes are indicated in Spiziapteryx along with a nedian one. The sternocoracoid processes jut upward, out, and forward and end bluntly. The sternocoracoid fossa is like that of the accipitrid. The costal margin has five to six distinct rib articulations separated by pneunatic fossae (fig, 767). The

Fig. L67. Lateral views of sterna of A. Faleo merieanus, B. Henpetothenes eaehinnans.

sternal plates are cupped anteriorly.for the viscera but atre flat posteriorly. The posterior margin is square; it may be complete, or have a pair of fenestrae or notches (fig. 168).

Fig. 168. Ventral views of posterior part of sterna of A. Ealco mesieanua, two specimens (a and b); B. Microhierat eaeruleseens.

The keel is thin; it is usually deepest along its anterior pi1lar. In outline it is roughly triangular with the anterior margin jutting slightly or distinctly forward and showing only a *** Evol. Theory 3:1-141 (October, f977) Part IV recelved September 9, L975; May 27 , I9TT 1977, The Unlverslty of Chlcago 9 2(202) M. JOLLIE

slight curve along the line of junction of the carina and lateral plates. It tapers to its terminus on the posterior margin. Henpetotheres and ui,eraatur show an uptilting of the anterior end of the keel like that of the accipitrid. The eupraeoraeoideus scar is regular in outline and extends fron one-half to three-fifths of the length of the sternum. Its ventral nargin is well separated fron that of the keel (thick pectoralis muscle). The outer tip of the coracoid is not braced by a ligament as in the other groups; therefore, at the point where the sternocoracoid fossa and the lateral line of the supraeo"aeoideue muscle intersect there is a distinctive, scarless, triangular area (fig. L67). The falconid sternum can be characterized as follows: anterior dorsal margin with a single rnedial dorsal process, or two lateral ones, inconpletely or slightly separated; coracoid sulci deep and narrow and extensively lapped; coracoid pads reduced to anteriorly facing contact lines I ventral margin of the sternocoracoid fossa lacking a ligament scar; a delinited triangular area between the sternocoracoid fossa and the sup?a- eonaeoideue scar. Subgroups cannot be characterized,

Sumrnaryand Conclus ions The sternum, like the shoulder area, shows such irregular variations from group to group that one wonders whether adaptive or phyletic features are involved. Since the forn of the thorax and- the pectoral musculature rnolds the overall shape of this bone, there is- little doubt that adaptation is a causitive mechanism. However, the consistency of overall body form within a family, and often throughout an order, gives this bone some phylogenetic value. Thus, the differences in myology, tendinal attachments, and liganental bindings, which differentiate the sterna of the falconiforms appear to correlate as well with ancestry as with adaptive- modification. The cathartid type is not exactly matched in any other bird, but certainly it resembles that of the pelecaniforms some ciconiiforms', and procellariiforns. If Tenatonnis is' assumed to be a highly modified cathartid, then the likeness with these other orders is strengthened. Fisher (l-945:728) was aware of the similarity of the sternum of Ieratorni.s and Diomedeabut overlooked the point-by-point agreement with Fregata. Not only the sterna but also the coracoids and, to a lesser extent, the entire skeletons of these genera are sinilar. Such similarity suggests some elenent of comlnonancestry rather than adaptive nodification for a particular style of flight. (No one would assume that they did, in fact, have a commonstyle of flight.) The sternum of Sagittanius is matched fairly well by that of Cari.anw,but comparison with procellariiforms, pelecaniforms and storks seems more likely. Comparison with the cathartid suggests comnon ancestry of much the same antiquity as that of the other orders mentioned. The Accipitrid sternum cannot be used for the identification of subgroups although there have been such attenpts in the past Ilere again it is necessary to consider the writings of Howard and FALCONIFORMES 3(2031

Loye Miller in regard to the Aegypiinae. Howard (1932:49) noted that, "The characters of the sternum of the nodern vultures, as distinguished from the eagles, are as follows: (1) broad ventral nanubrial spine; (2) low carina; (3) short and broad sternocor- acoidal inpression. " Figure L69 shows a view of the sternum of Neogyps and several of the Aegypiinae. The coracoid sulci are widely

Fig. 169. Dorsal views of anterior part of sterna of A. GypaEtus banb.atus, B. Aegypius monaehus, C. GApe eop"otheres, D: Neogyps e"Tans (2 specimens). separated on the nanubrial spine in Neogyps (range from well separated to in contact) and GypaEtus; those in Torgos and Aegypius are nearly in contact; Tnigonoeeps, Saneogyps, Neophnon, Necroeyntes show incipient lapping while in GAps, Peeudogyps, and Gypohiefat they are distinctly lapped. ?enathopius agrees with Tongoe and Aegypius in this detail. Most accipitrids show well lapped sulci, but in some specimens of Buteo and Leptodon, they are only slightly lapped. In fossil sterna of AquiLa it was noted that some showed no lapping and every transitional phase to that found in the modern species. On the bases of breadth ot manu- brial spine and small size of the anterior inner process on the 4 ( 204) M. JOLLIE

sternocoracoid process Leptodon compares well with the aegypiin. is to evaluate. The relative depth-Gypohierar of the carina difficult That of Neophnon or can be included in the range of variation fbund anong-the typical accipitrids, whereas that of AquiLa is definitely lower than in the majolity of genera. Neogype is like Aquila in this respect and it resembles that genus- in alnost every way. Howardts last point on the relative size of the sternocoracoid irnpression ov-erlooks the very smal1 fossa of Pandion and the overlapping series fron AquiLa through Necnosyrtes"- Gypohienar, to that of the Large aegypiins. Neophnon,- iira ileophnontope A more charatteristic feature of the sternum of the Aegypiinae is the forn of the yup?aeo?aeoideue scar. This is fo['ei along its posterior lateral nargin, and_the-ventral (nedial) margin liei close to the edgq of the keel. The shape of the scai in Tongoe is deternined in part by the pectonaLi,s muscle. Neophron tends toward the aegypiin outline, whereas ,Neo-phnontops and'Cypohierac are sinilar to the typical-accipitrid; i.e., they in this lack ihe lobe. Teyathopius-sagittanius. agrees with the aegypiins character as does also A type of sternum, limited to the subfamily Aegypiinae, cannot be characterized by arLyone feature or conbination of features, unless the species content of that subfanily is altered. The latter step would necessitate removal of Neogyps, The falconib sternum superficially resembles that of the accipitridae, and it does not find a better comparison elsewhere. However, this does not require a commonance-stry for these two faniliei as the conparison is still poor. The sternun of these two families, thusr-gives no clue to their ancestty, at least at this reading. OSTEOLOGY..MISCELLANEOUSSKELETAL PARTS

Under this title the vertebral column, pelvis, humerus, ulna, manus, tibiotarsus, fibula, and foot will be considered, more by discuising features reported in the literature than in atternpting to delineite and evaluate new ones. Since any of the above can be described rather briefly the types characterizing the families of the falconiforms will be discussed together.

Vertebral Column According to Beddard (1898:1L5), the total number of vertebrae in the spinal column of birds varies from 39 to 64, counting the pygostyle as one (based upon-Giebe1, 1866). According to- Parkbi.-(1Sb0:17), 'fMore than half of the tuelue thousand s-pecies known hiver otr an average, only forty vertebrae, even in the fhe African oslrich h'as-fifly-sia, and the-common swan, eirty-thnee."rUtio. NumericaL law is not very strict in birds; yet thosl six thousand species just mentioned have in a great majority just founteen cervici1s....'rueloe also is a conmon number for the -aoian-sae?um in the small high-type birds; and from ten to twelve vertebra in the caudal region is as a rule the number in in the falconiforms the embryo." The total number of vertebrae ' FALCONIFORMES 5(205) not including the pygostyle, varies from 39-41- among the cathartids, from 37-41 (usually 40) arnongthe accipitrids, 40 in Sagittarius,- and 39-41 (usually 40) among the falconids. Ordinarily the vertebral column is divided into a series of artificial segments: cervical, thoracic, lumbar, sacral, and caudal (fig. 170). The cervical vertebrae are those of the neck region, which extend posteriorly to (but not including) the vertebra bearing the first conplete rib. Thoracic vertebrae begin with the first complete rib and extend posteriorly to that vertebra bearing the last (floating or vestigial) rib. The lunbars bear no ribs and continue the series to the sacrals, which purportedly support the pelvic girdle. The caudals extend posteriorly flom the sacrals (but may be indistinguishable fron th-en) to the tip of the tail, the terminal caudal vertebrae fuse into a pygostyle. Part of the thoracic series, the entire lunbar, and sacral series and part of the caudal vertebrae are fused into the synsacrum (Table L2, fig. L73). Infornation on the brachial and pelvic plexuses supplies an approach to comparisons of the segments of the body i1 th-e diffLient falconiforns. Figure L7L shows diagrammatically the distribution of the brachial roots in relation to the cervical and thoracic vertebrae as based on ribs; Figure L72 indicates the roots of the pelvic plexus. Homologizing nerve roots necessitates criteria such as thickness, main apparent pathway of fibers, and detailed sinilarities between reLated forms. Using these criteria, comparable roots have been blackened in the figures and the resulting segments of the body have been summarized in Table 15. Sagittarius, the Accipitridae, and the Falconidae are found t6 agree in having L4 vertebrae in front of the brachial root and nine between this root and the pelvic one--the exceptions among the accipitrids being highfy specialized forms in which these nunbers are nodified. One can conclude that the nunber of cervicals, based on the position of the brachial nerves, is 14. Modification of the nunber of cervicals involves displacenent of the pectoral appendage posteriorly (a relatively loni;er neck and shorter body) or an increase or decrease in the nunber of cervical vertebrae produced during enbryonic segmentation (eee Goodrich, 1930:L37-L42); the former is the more probable.- Among the accipitrids there are one to three incomplete cervical iiUs; the usual number is two. In most species the first cervical rib is a short, triangular segment with both tubercular and capitular articulations. Occasionally this vestige may be fused to the vertebra and thus removed fron the list of ribs; such may be the case in Gymnogenysand ELanus. The second cervical rib is well developed, tapering distaLLy to a point. In Gymnogenysthe second rib (assuming the_above) is a small, triangular nubbin like the first cervical of other accipitrids. itre first thoracic rib of several specimens of fetinia was inconplete. It is possible that other specimens may have this rib complete. The cathartids have five or six complete ribs, plus an 6(206) M. JOLLIE

Fig. 170. Segmentation of the prepelvic portion of the verte- trrif column in A. Cathantes auna, B. Gymnogypa eaLifornianus, C. Sagittantus serpentanius, D. Aquila ehnysaAtus (Buteo, Hali'aeetus, E. Gyps eop?o theres F . Tongos traeheLiotus BubaneLLus) , , @.egypi'us' Pitheeophaga), G. ELanoiidea fonfieatus (Neophnon), H. retinia misisippiensis, I. Chondnohienas uneinabus, J. Pandton haLiaetus (Pennib-, Harpia) , K. Herpetothenes eaehinnans (Miet'astun) , L. icrae'idea beiigona (PoLybot'us, IvtiLoago, Mierohieras, most species o I' ITaLeo), I{. tr'aleo eolumbarius , N. FaLeo eoLunbarius (lWZ 15631) . FALCONIFORMES 7(207 )

Table L2. The number of segments in each region of the vertebral column for selected species of each of the basic types. In some cases there is a rarlge of variation, the nore common nunber is underl ines

?n r-{ Fi Fi Ul (d ddu+) h UU.r{tr u .t-J () .Fl .r{ (.) O 'r{ o) .Pq FrE tn t{ t{ O t{tr tr h 5 5 t-{'-{ O O F{ go 0f \ o OUtr O<\l r< .o .o (/l € o €t/) S'-J lr .o€ ed E F.O E.O E Ei6 EPo E5 +) A 5.'i O.'-r 5 tr 5.a 5f{+') t(d O tr H t{ u t{ tr.,r trh trcd(n trU +r Cathartid Cathartes aura 15 2 4 t4-L5 _,t-2 J 5-6 39 ) -6 Cathar te s 1_5 L 4 L4- 39 urubitinga Vultur gryphus L7 2 3 2-3 5-6 5-6 Gynnogyps 18 2 3 14-L5 Z-s 5-6 5-o 4L cal i forni anus Sagittariid Sagi t tarius 1_5 15-16 2-3 6 5-6 39 serpentarius Accipitr id Pandion haliaetusl5 4 L5 z 6 6 40 Aviceda L4 4 15 738 cucul0ides Pernis apivorus 1_5 7 4 2 6 Chondrohierax L4 2 J 1"5 2-3 4 737 uncinatus Leptodon 13 13 ) 737 palliatus Elanoides L4 L4 2 639 forf icatus Elanus 14 13 t 739 leucurus Ictinia 15 L5 L-2 739 mis is ippiens i s Rostrhamus T4 2 6-7 sociabilis Gymnogenys 1.4 2 typicus Machaerhamphus L4 L4 2 40 alcinus Milvus L4 5 14 7 40 migrans Hal iaee tus L4 4-5 14-15 ? 40 leucocephalus Gypohierax 15 4-5 1 angolensis 8(2oB) M. JOLLIE

G1'PaEtus 15 2 4 L4 2 6 5-7 barbatus Neophron L4 2 4 2-3 6 perenopterus Haematornis 14 z 5 L4 z 7 740 cheela Terathopius 15 2 4 16 L-3 '6 6-7 ecaudatus Torgus, AegYPius 15 2 3 15-16 3 6640 Gyps, PseudogYPsL72 5 15 2 5 5-6 4L Necrosyrtes 15 2 4142 6 5-6 monachus Harpia harpyj a 1-5 2 4 L5 z 6 74L Pithecophaga L4 2 4 3 7 7 j efferyi Aquilins and t4 2 T4 40 Buteonins Falconid Herpetotheres L5 3 4 L4 40 cachinnans Mi cras tur 1"5 3 4 L3 39 semi torquatus Milvago t5 3 5 chimachima Polyborus l-5 3 5 t_3 40 cheriway Microhierax 15 3 5 1 S7 caerule s cens Falco L5 5 5 L2-t3 r 6 7 39-40 mexicanus Falco columbariusLs 3 5 L-2 67 Z 5 T-z 77 Falco columbariusl-4 *tr *******fr*******?t* ****T*** **tttr*tr additional vestigial rib posteriorly. _ The number of thoracic ribs not included itr itt" synsrttur ranges from three in Gymnogyps and Vultur to four in Cathantes. The last genus has two full synsacral ribs and the former genera have two or three. A11 may have an additional vestigial rib. Sagittart4s has two cervical and five or six complete thoracic ribs; tirere is a vestigial seventh rib in some specimens ' Two ot the normal .orpi"*"nt 6f thoracic ribs, plus any vestigial seventh, articulate witir the synsacrum. Sagittarius resemble the large aegypiins (and Cariama) in its thoracic structure ' Usualiy the accipitrid has seven complete thoracic ribs, two of whiCh articulate with vertebrae fused into the synsacrum. Occasionally there is an eighth (posterior) ribr and it may be complete or vestigial. The-extreme is reached in ButeogaLLus (MYZ 85546), whic[ has a ninth, vestigial costal segment and ir"onnllus- (WZ 85549), which has both vertebral and costal segments of that rib. The number of ribs is reduced among th9 aegypiins, certain kites, Pandion and Harpia (s?n Table LZ) . lg- auEtion reaches its extreme in GAps, which has but five full ribs; two of thesc articulate with the synsacrum. FALCONIFORMES 9 (209)

Fig. L7L. Roots of the brachial plexus of A. Cathartes aura, B. Conagyps atratus, C, Sagittarius se!'pentatius, D. Auieeda euberistata, E. Buteo jamaieens'ts, F. Torgos t.raeheLiotus, G. Pandion haliaetue , H. PoLyborus eheniuay , I. Ealeo sparuerius. Key root shown in black.

Fig. I72. Roots of the pelvic plexus of A. Cathartes aura, B. Sagittarius serpentanius, C. Buteo iamaieensis, D. ?orgos tnaeheliotus, E. Pandion haliaetus, F. Fale:o sparuenius. Key root shown in black 10(210) M. JOLLIE

Table 13. The number of segments in each region of the column for selected species of each of the basic, types, based on the position of the main brachial and femoral nerves. In one case there is a range of variation, the more commonnumber is under- 1ined. number of verte- number of verte- number of vertebrae anterior to brae between nain brae posterior main brachial brachial and nain to main femoral nerve femoral nerve nerve Cathart id Cathartes aura 15 8 16-t7 Coragyps atratus 15 8 16- Sagi ttari id Sagittarius 'J,4 16 Acc ipi tr id Pandion 14 L7 Aviceda subcristata 13 16 Chondrohi erax uncinatus L4 L6 Buteo j arnaicens is 14 L7 Hal i aee tus leucocephalus L4 9 L7 Necrosyrtes L5 8 I7 Torgos 15 8 I7 Falconid Falco sp arve r ius L4 L7 Polyborus cheriway L4 9 L7 rtrtrttt***tt** rttctt*rttt rt*?t rttt****:ttrrs**rt2t The falconids have three cervical ribs (actually two true cervicals and one thoracic as indicated in Table 13). The total number of complete ribs is generally six or seven. Eight ribs occur in a few specimens of Falco, but in one specimen of Faleo merieanus a ninth, vestigal rib was noted. Sushkin (1905:38) pointed out that in the falconid the gap between the transverse processes of the thoracic vertebrae is equal to or less than the width of the processes, whe_reas in the aicipitrid the space is more than double the width of the pro- cessbs. Measurements, summarized in Table L4, indicate that his views are largely true. The particular Specirnen of Mierastut' cited.has the widest gap of the falconids examined, whereas SpizaUtus ornatus is peculiar in the great width of its transverse processes. Further, the falconids, with !h" exception of nenpetot'lieres and Mi.erastur', have four (or f ive) of the thoracic vertebrae fused. There is some variation as to which vertebrae are involved and their relationship to the pelvis (fie. 170). funong the accipitrids and falconids there are usually seven free caudals (exclusive of the pygostyle); occasionally five or FALCONIFORMES 11(211)

Table L4. Relative widths of the transverse processes of thoracic vertebrae in the accipitrid and falconid. ab a/b width of transverse gap width between process in mn. processes in mm. Herpeto theres cachinnans 3.5 4.0 - 4.5 77.7 - 87.s Micrastur semitorquatus 3. 8 5.0 - s.2 73.0 - 76.0 SpizaEtus ornatus 4.3 6.6 - 6.8 63.2 - 65.0 Aquila chrysa6tos 4.3 10.0 -1.2.0 35.8 - 43.0 Heterospi za meridionalis 2.7 7.2 - 7.7 35.0 - 37.s **************tr*tt *rt:t** ****r.r.fr**** six ane found or rarely eight. The number of free caudals varies because of the inclusion of one in the syncacrum or fusion with the pygostyle. This is particularly marked in the cathartids wheril'if there are L4 segments in the synsacrun' there will be sili free caudals (one nay be fused to the pygostyle); if there are fifteen synsacrals, there will be five caudals. Sagi'ttarius is like the cathartid in the number of free caudals. The pygostyle is nade up usually of five units, the last three of ithich cannot be detected in the adult bone; the first two units nay be clearly marked. A caudal vertebra nay be to this structure. partially- or almost completely fused Assbciated with the pygostyle in most falconids is an accessory ossification at tha-insertion of the depreseo? eoecygis muscles (see Table 18). According to Sushkin (1905:74) t\is ventral. coccygeal sesamoid is found in all of the falconids except the pblyborinae' Polihieracinae and-"Harpie" l? FaLeo nouabeeeLandiaef, Sushkin believed that this sesamoid is correlated with the strong taiL musculature needed for "steering" among the falconids. Pycraft (1910:339) viewed it as a brace for ltte tail feathers when they are used to check the rush of flight. Amongthe accipitrids and cathartids the amount of coniective tilsue here is much less than in the falconids and no ossification takes place. The development of the free caudal vertebrae, and pygo:tyl-e corresponds with the use of the tail. Minimal developnent is observed- in Terathopius where the tail is short and sna1l. Beddard (189-8:L15) concluded that, "The nunber of vertebrae...is not of the faintest use for the systenatic arrangement of existing formsr" however, the cervicaL or caudal nunber have some valuel-although it is probably never a sharp or absolute one. In the case of the falconiforms the falconid type is differentiated on the basis of the number of "cervicals" and the fusion of thoracic vertebrae anterior to the pelvis. The cathartid is marked by the increased number of cervicals, which is rnatched only in aberrant genera of the accipitrid. The sagittariid is like the cathartid in the apparent number of ceivicals but like the accipitrid and falconid in the true number. It differs from the latter in the number of "cervical" ribs. L2(2r2) M. JOLLIE

Pelvis The pelvis is formed by the fusion of a number of vertebrae, the synsatrun, with the bones of the-pelvic girdle. .The segnents of th'e synsacrun are'exceedingly difficult to count in some adult birds, but the nunber of spinal nerves observed in the flesh nakes-such a count simple (fig. L73). In the natal or early nestling the segments are well marked, fusion occurring-1?!9 in Cttu groith periSd. The pelvis nay undergo a degree of folding .luriig developnent, which tends to obliterate any ventral trace of ceitain segments. The number of vertebrae in the synsqqrYn in the cathariids is usually L4, but a fifteenth can be added fron the caudal series. The accipitrids usually have L4, some kites have L3, and some of the aegypiins and Pandion have 15. iagittiriue his 15 or !6, depending-on the inclusion of a caudal. Rm6ngthe falconids, Faleo and Mierohi.erao have LZ (Faleo MVZ 54386, appears to have 11).or \1; tI" Polyborinae h;;;"joi7"ri.ue, i3, n"opetotherZ" is alone in having L4. Variation in number in ^id the cathartids and the aegypiins involves not only the inclusion of a caudal but also fbiward extension of the ilia to add a thoracic. Norrnally the anterior nargin of the iliun'is even with the anterior face' of the first synsacraL, but in these vultures this margin extends to the niddle of the added Verte- bia, which nay be"firnly fused into the synsacrun or remain free. The tyrrt".t.,rn is firnctionally equivalent to the sacrum of the nannal, although it is made ub of a much greater.number of units, some of whiEh bear ribs. ltre attenpt to-distinguish parts of this series as thoracics, lurnbars, "sacralsr" and caudals can be viewed only as artificial . In the developnelt..of this structure theie is no clue as to the presence of "true" or sacrals. The entire structure is involved in support "primary" included,- o'f the ielvic girdle with the exceptt4n only of partially unfused, thoracic or caudal vertebrae. Vi6wed ventrally the segments of the :Insacrum, show varying development of the transverse processes; this is related in-part io itt"^kidney. The strongest transverse supports lie anterior secondary to the acetaLulum, and th; kidney, with -supports-be-- . hind. The latteri usually identified in parq as the sacraLs, lie stlong embeddedin the kidney. Posterior to the kidney -are -sup- p;i;-for ih" potteribr margin of the innominate bone of either iiae. The s.rpports in the iegio! of the acetabulum may show asynmetric development or reduction. The number oi ribs (i.e., thoracic vertebrae) articulating with the synsacrum varies; it is not uncommonin any species to find one more than the usual number. In the devefoping pelvis there is a fenestra between each of the transverse processes. Anteriorly these tend to become ioverea by medial extension of the ilia; po:teriorly they nay be closed by'bone growing inward from the margin. Closure appears- to result from lhe downward rotation of the posterior part of the innominate bones and the consequent shift in origin of the Leua- lii .o"oygis and Let;ator eaudae muscles. _ T!t" strongly fenestrated peivis h;; a straighter dorsal outline of the innoninate bones ind the anterior eids of these muscles reach forward beneath the FALCONIFORMES 13(213)

A B a

Fig. I73. Ventral views of pelves of A. Cathartes auna, B. Sagittar'ius serpentaniua, C. Buteo iamaieens'ie, D. Pandion haliaetus, H. Faleo meti.eanua. Arrow indicates level of center of acetabulun. 14( 21q) M. JOLLIE

anterior plates of tJte ilia (Conagype,and Elanus--fig . L74). The i'prinitive pelvis,r' that with op'enilgs- palticulally in the posterior half, is found in each group- of the falconiforms. (Machaerhamphue)to smalL (Torg-os Fenestrae range fron large ' Auieed,a, f eptSdon" Chondiohieraa, rctinia, Gymnog??A8, and Geranoep'i.za)anong the accipitrids , lLthough usually they- are lacking' as most slecies havb a strong- downward curve of the axis. Far poiterior paiis of fenestrae m-aybe found,in pTl9tically any- tpu.i"r as a cbrrelate of age or the degree of ossification reach- eit in the individual. Fenestrae are usually lacking in the falconid but appear in PoLyborus, and some sPe,cies-of Faleo. According'to Fisher (i946:641), "The height of thg (9psacral) vertebrae in itre Cathartidae dininishes rapidly posteriorly. The decrease in the height is nuch more gradual in the hawks and the reduction shown in Cathartes, eagles. " Chondrohi6nao-equals whEreas the synsacrun of sorneof the aegypiins ot Sagittarius may exceed it (fig. 175). ' Lying lateral to the synsacrum, ald showing varying de-grees of fusion-with it, are the innominate bones formed by the fusion of the ilia, ischia, and pubes. The ilium lies dorsaLLy and. extends the entire length oi the pelvis. Anteriorly it is a thin ipoo"-shaped plate fling at_an angle.-- This anterior hollow is ii" bf oiiein of tf,e gtuteus-med.iue nuscle, while the dorsal crest"ru" narks the-origin of the tensor faseia Latae (see section on ;t;i;gt for muscle iarnes_). Posteriorly the dorsal crest has been drawn out as the line of-origin of the i.Liofi'bularis; it overhangs the ilioischiac fossa from which the obturator internus arises. This fossa is nargined anterodorsalll by the antitroch'learis mass and anteriorly by"the colurnn of the fusba acetabular portions of the ilium and ischium. In the anterior half of the pelvis the ilia meet for a short distance along the rnidline above tne synsacrun. Viewed laterally the dorsal onfline in this region of meeting is fl1t, but in the AquiLa and other_ powerfil forms (Spiza7tus) is distinctly ;;;"; .it in ircfrea ifig. 176). Arching of this crest reaches its extreme iijitt"iiu|, Spi'zaEtue, aid Aterastun; in the last two.genera the foim of the- cr'est nay be correlated with a narrow pelvis' In iigitto"i."" it appears to be associated with the cursorial habit. eoiteriorfy ttre ibrsal nargin of the ilium angles-downward. This slope is strongest in the lowerful predators and least in the kites and scavengers. Posteiiorly, the ilium may not b-e fused to the sln:acrum. The cathartids (wittr ttre exception of VuLtur) ?nd accipitrids line of junction (fig.^ have a distinct crack along this \7a): - signttarius shows conpletg fusion here as do also the falconids. Viewed from abov'e, the lateral outline of the posterior iliac crest varies due perhaps to the influence of the i.LiofibuL.arts- muscle. Anong th; accipitrids the lateral margin J.s distinctly S-shaped. It curves over the ilioischiac fossa, abruptly.ltt 1! _, the pbsterior tip of the iliun,.and then out slightly as the margLn of tire ischiun. Lateral extension of the nargin is greatest in Pand.i,on,which has an extremely broad pelvis; it is least among some of the Aegypiinae (especibfty neophron) and Ln Leptodon and FALCONIFORMES 15(2r5)

G Fig. L74. Dorsal views of pelves of A. Cathartes auna, B. Ggnnogype eali forni anus , C. Sagittanius senpentarius, D. Neophron pe?enoptenue, E. ELanus Leueuz,ue, F. Buteo janaieensis, G. Pandi,on 'haliaetue , H. Mieraetur semitorquatus , I. FaLeo mesieanua. 16(216) M. JOLLIE

A. aura, B. Fig. L75. Lateral views of pelvet o{ .Cathartes Buteo iamaiceneis, D. Leptodon SiTlttanius serpentarius, -C-. poiliotu", E. pandi,on haliaetus , F. FaLeo mes'Leanus.

Chondrohierax. In these accipitrids the nargins are- nearly ;;;;ighi and paralle1 as in tie cathartids. Anong.the falconids itt" nirgins slightly outcurved with the exception of Het,pe'to-theons,"iu wherE th'ey are para11e1. Sagi.ttar:)us is unique i;- lh;|, af ttt,iugtt these inargini are like those of the aegypiins , fossae. they----' extend far out over the ilioischiac The depth of the ilioischiac fossa is variable; it is deepest in sone accipitrids where the posterior dorsal crest of the ilium has been diawn out. It is 16ast developed in lhe cathar- ;iA;; aegypiins, kites, and falconids. S_geittatius is peculiar in the oveihanging iliac crests (fig. L77). FALCONIFORMES rT(?-rT )

Fig. L76. Lateral views of pelves of A. Sagittarius serpentarius B. Spiza7tus onnatue, C. Mierastur semitonquatus, showing arching of dorsal iliac cres t .

Dtr

Fig. 177. Posterior views of pelves of A. Cathartes auna, B. Sajnitaniue se?pentarius, C. Buteo ianaicensis, D. Pandion haliaetue, E. Falco mexieanua,

The ischium arises as a stout backward-projecting pillar at the acetabular circlet and then spreads out as a thin plate, which fuses with the iliun behind the large round ilioischiac (ilio- ischiadic or ischiadic) fenestra. Ventrally the ischiun lies in contact with or close to the pubis. In most accipitrids, and in some specinens of Sagittarius, the fused plate of the iliurn and ischiuln, &s viewed lbterally, tapers to a rounded point, which rB(218) M. JOLLIE

lies next to the pubis (fig. 175). In some specimens of sagittarius, some of the large aegypiins and in Leptodon, Chondrohierau, and Pandion the posterior margin has a distinct notch. This margin is truncated in the falconid, aTthough certain members of Faleo may show a wedged form sirnilar to that of the accipitrid. The cathartids stand apart from the others by virtue of a deep irregular notch here. The pubis is rnuch reduced and appears as a long, ventro- posterior projecting splint attached below the acetabulum. The pubis nay be fused or closely bound with the ischium behind the ovate obturator forarnen, or these bones nay be separated by a puboischiadic gap or fissure (cathartids, Sagi,ttarius" Microhierar) . In some falconiforms a segment of the pubis lying behind the obturator fenestra is lost. Pandi.on is unique in having an extremely widened region that contacts the posterior ventral angle of the ischium, although a somewhat broadened pubis occurs in Herpetotheres, GAps, and Sagittarius. The posterior tip of the pubis in Coragyps is spatulate. The sagittariid pelvis can be characterized, although it is matched in nany details by that of Cariama. It resembles the typical aegypiin in some respects, differing in the arching of the anterior dorsal iliac margins and the overhanging of the posterior margin (or medial movement of ischial and pubic components). The number of vertebral segments exceeds that in the other fanilies (not always the case). The cathartid series is continuous with the accipitrid but tends to differ in the following ways: weak transverse processes at the posterior limit of the acetabulum, pouched posterior margin of the ilioischiac fenestra (which is either pneurnatic or unpouched in the accipitrid), heavy rectalinear form with straight lateral margins to the posterior part of the iliurn, and a distinct deep notch along the posterior margin at the junction of the ilium and ischiurn. The pelvis of the cathartid is most closely approached by those of some kites and aegypiins. The accioitrid series is discontinuous and varied. The cathartid-1ike pelvis of kites and aegypiins, the narrow pelvis with sweeping lateral curves of the more powerful rnembers, and the broac perforated type of Pandion, Elanus, and Maehaerhamphus tend to con- fuse rather than clarify. The most peculiar pelvis is that of Pandion, which can be included in the accipitrid series only in the most general wayi it is as distinctive as the sagittariid or falcon- :A The falconid pelvis resembles the accipitrid, differing only in the type of postacetabular transverse processes and the fusion of these with the postericir extension of the ilium. The number of segments is generally less than that of the accipitrid. The forn of the pelvis appears to be correlated in some cases with adaptive modifications. Those forms which are more cursorial (or use the legs in diving) have relatively narrow pelves of rectalinear outline (lacking strong curves). T'his shape also can be partly correlated with powerful, grasping feet--the preacetabular portion is proportionally longer and narrower, whereas the FALCONIFORMES 19(219) posterior margins are broad and strongly incurved. Those which are less cursorial have very broad pelves, even though the legs may be powerful as ;n Pandion (see Pycraft, 1910:390-392. The tinanou pelvis does not fit into the above; nor does the sinilarity of the pigeon pelvis to that of the tinamou help in understanding the variations of this bone. Since these correlations are not very exact, the pelvis can be of taxonomic value only at the extremes of the taxonomic hierarchies. At one extrene, it is possible that the separation, or near separation of the iliurn, from the ischiun identifies the palaeognaths (palaeopelves) in contrast to the neognaths (neopelves) in which these bones are always broadly fused. At the other extreme, closely related genera or species should share similar pelves. At the fanily or order level only suggestive similarities supporting interrelationship are wanting. Humerus It has been suggested that the humerus of the bird might be of taxononic value (Sirnpson, 1946) , but as yet a study of its variation has not been made (McDowell, in Mayr and Amadon, 195L:9, has based broad taxonomic conclusions on it, see also Berger, 1957:240, 266-267). Some fossils have been described on the basis of this bone. The type specinen of the Pliocene cathartid, Saneoramphus kerneneie, is a somewhat crushed fragment of a humerus (Miller, 1931). The specimen certainly resembles this bone of the modern species, S. papa., particularly in the possession of a prominent rounded tubeicie at the distal end of the deltoid crest. In the living species this tubercle is the point of insertion of a strong tendon of the superficial layer of the peetonaLis muscle. A sirnilar tubercle is found on the humerus of Chauna ehauaria. The shape of the humerus is fairly consistent in each of the falconiform groups, but it is unlikely that this bone could be used in an effective characterization of them. Pandi,on is unique among the accipitrids in that there is a deep pit on the anterior aspect of the bone between the distal condyles. The functional significance of this pit is unknown; the braehiaLis muscle arises from it in the same fashion as in the others.

Ulna Howard (L932) attempted to characterize the ulna of the Aegypiinae and to a1ly Neogyps with it. Regarding Neogyps, it is pointed out (p. 53) that, "This similarity is noted particularly in the narrowing of the palmar side of the shaft near the proximal end, and in the relatively short external condyle at the distal tt* end. A comparison of the ulnae of the various genera of the Aegypiinae reveals no cogent features. The measurements given in Table L5 differ from those made by Howard but indicate the same * * * * * * * * * * * * * * * rt * * * ?t * 2t tt * * * tr * * * * * * * Howardrs terninology differs from that of Fisher (1946); in the figures and text "palmer" and "anconal" are anterior and posterior aspects, "external" and "internal" are dorsal and ventral. 20(220) M. JOLLIE

0) (>oc\ dt BUIn +J F{ Fl f\ f\ Fl f'* Ot ri Ml ln C) 'o r{lEueT/tlrEueI,(Pog r,o rn

ri'O € ef,"Jrns '1rB SOcd .+OOC)!AFl t{.P Jo r{lEua1/e1.(puoc .no@|-rrrr1 opc) ssoI3? OTOOO)Olr-{OFIOOIF{ 3ou 'PUe IEISTp F{ Fl t-t F{ t-{ F{ t-{ +rd(d tf,eds3 TssroP tIIPTM o qqEcl og |/)+J (n(tl tr r.'l-{ eI{Puof, o >o .'{ FO ssoJf,B 'pua IBIsTP 14@C)C)GtO(>F._t.f1 O€ +rotr qlPTr lcadsB IBsroP rn tn F{ (d .O'rf l-- \O Ot c{ F\ c{ Nl A F.l b0 /'totd 'luo^ UIoJJ ef,BJ @ootot@otorototc)ol SOtr FF u& c! -rns 'tre sso.lf,? qlPTlt .'-.1 {J E +.) F. +l F{ (d00 euln d.q li sff-f\€@@F,tl- od +r q) qlEue1/e1(puor €€t/)t-lFltic{€tf,€\O p bo+r Jo aa oCtr ssoJt€ 'Pua IelsTp F-- \O \O \O t- F.. \O \O @ l\ co '.{oo qIPTM EFI l3adsB I€sroP C) € ot tr,q {J (0+J o pua 'xo"rd ssesord F{OP Jo CdP 'rop qEnorql {ldep E(6 .rf(0h 1sel€erE /'tte Jo lEIJ x trF '3ord esBq tn rn oo \o \o c{ o =t Fi or }a O'-l O T"sJoP JO F.- F\ D- F\ F\ f\ F\ f\ f\ \O F\ tr tP 'lredse roTreluB I{lpTM Ah oo ot > SP PU +r €uIn ||oO 'Fl Ln rO F. f.) 14 4) r\ od Jo r{lEueI/'tJ€ J:o l"TJ 3 Gf6rnoc{r+o

O +)l| .r-{ O ln 5(n k(H .r{ q d o'd ti O O Or-l O ln .r''l u) ?n trc) +r(Do gd t/) .,l.AA Xo..r l^ 5 .o p{+J5 o+J 5 5.-t+JCdHH 5 +r Ln(^k C O k'< t, U.'-{ tn.d n +.l triti F{ O h (6 t{ H (d F-{ .r{ o tr XU 5 o A3 u 5 (d o'O o'-{ oFt o o.''{ O (d f{.rl t{ O t/) ) H..r.'{ d+r.Ot 5.rr O+r O O >€rr ,C li O tr r{ O U p{tr:O F p cd'4 bod O.F{.F{ OF{ o(6k (d (d p.P r{ A0) p.O ${ O ln 5 bou >.o d u o tr Et tr d O OiU E Ar+{.r{ O @E p.,o t{ O Atld O U h cd AQ !Ep(6otr oox'{oxoxdoo F{u(d ZZgFr FALCONIFORMES 2t(22t)

t\ (D \o Or(rlrn@u)r-C)€ ....1 (\ f\ t\ cl (\l F{ Ml \o ra \o or r\ \o '\o -t\o o ol v1

(\l :f O !4 Ml ln t-{ Ml

Fc'tOMtF{OMl€ OoOoOOOOI r{ F{ F{ ti r{ F{

-t14OF-r\C-OGl F\€O€€Gl\O@ Orr{O€OlOlCn€ r{ F{

(\6N\OrnC\ral\f4u1 u1 r\-f€-f€€oria

f\cltrAMlf\FO

Gr€ra€rn}.)olra rntnOGl(\€OlF\ " €f\.C\Ol-\O\OF

ar't rn 14 !.) Gl f\ O OGlGlt-tF-OiAGtO

GTOOF{(flOlOlOO FJ F{ r-{ Fl Fl Fl

o .r{ l/) qa XtrF{ rdtnoGl(,)l,l t{rn A .r{ q .€dt O O 5 P A 'F{ 5 t4 AOO+r P .r{ +J t,) (d gt{.r{ 'F{ = (D 6:(D :O *(dOU'o .r{OF{ tr+rU,4td td o €'o.F{ cl tn +r () o o o Ax !/| . x cr q t{.r,ttr{t+a.r.t.r{.F,to{O u) O U O.c cd >\co d A(ll € u o t{ c ln Ao 5€ d =| x+r-t t.dd >t{ c tr c o.r{.'.{.F{ o u 5.'{ o.q u..{t.d 3 boh ij dt| +) E U U h.dr

relationships of which she speaks. From these data and the bones themselves, one can say that the ulnae of the Aegypiinae are usually very large, fairly straight, and relatively longer as conpared with trunk .length (from anterior end of first thoracic vertebra to anterior margin of acetabulum) than those of other accipitrids. However, Terathopiue nust be included in the Aegypiinae, whereas Gypohie?a& does not conform.' AquiLa audas has the longest ulnae of the typical accipitrid group and approaches the Aegypiinae, just as does the fossil genus Neogyps. Assuming that the trunk of Neogyps was approximately of the same length as that of the golden eagle, but the wing was L04 longer (as based on a reconstruction at the Los Angeles County Museum) the ratio of body length to ulna length would be fron 55-61% or the same as that of Aquila audas, The ulnae of Neenosyntes and ?enathopius show a distal tapering, which is not shared with the other genera of the Aegypiinae. Trigonoeeps, Aegypius, and SareogApl have a proxinal pneumatic pit on the ventral (inner side); this is developed in GAps and Neenosyrtes. Another fossa, at the base of the proxinal dorsal process (= the external condylar porcess), is found in GAps, Aegypius, Saneogyps" Trigonoeeps' and Neenosy?tes. Ierathopius agrees with AquiLa in showing only a shallow im: pression. The shape of the external condylar process tends to be more squared and less drawn forward anong the Aegypiinae although Neeroe[rtes and Terathopius bridge anY gap that night exist here. The form of the ridge that runs distally from this process is distinct in the larger aegypiins, but again the group lacks agreement- among all of its members. The distal end of the ulna of the Aegypiinae shows a tendency for pneumatic formina, just as does the proxinal 9nd, although here too the variability is too great to serve for characterization of the group. 0n every count the Aegypiinae overlap the typical accipitrid. The ulnae of the kites show a genuine resemblance to those of the large vultures. The ulna of neogyps may resenble that of some members of the Aegypiinae, but this is not evidence of relationship because of the heterogeneity of that subfamily and it,s intergradation with the other accipitrids. The resemblance of Neogyps to one of the large vultures is no better than to the eagles. Palaeoborus roeatue was described by A. Miller (1944) from a fragment of the ulna and was included in the subfamily Aegypiinae.

Manus There is a small bone in the wrist of the accipitrid which articulates with the "radial" (fig. 178). This bone was first illustrated by Milne-Edwards (1867-71) in a drawing of the wing nyology of Aquila (rlroa7tus) audas. Mivart (1873) also used this illustration and cal1ed the element a sesamoid bone. Alix (1874b) identified it as the "os prominens" or "epicarpium" in the kestrel; in L881 Shufeldt used this name for a bone in the wrist of Cireus eAaneus. Shufeldt also noted its occurrence in several species of hawks, including Pandion. Lucas (1882) FALCONIFORMES 23(223)

iiff ;l')

Fig. 178. X-ray of left wing of specimen of l,ecipitev, coopez,ir) (ulna broken distally) showing os prorninens on anterior aspect of wrist (arrow). described and figured an os pz,ominens tn Bubo ui,r.gintanus and noted its absence tn Strir. He noted its presence tn SareogAps and its absence in PoLybot,us and ailuago. Shufeldt (1909) agreed with Lucas in all details and, on page I29, suggested its presence be used as a character of the superfamily Falconoidea. After this brief flurry nothing more was said concerning this bone until 1959 when Holmgren commented that it answered the description of the true radial rather than the fused centralia now identified by that name. Shufeldt's (1881) description of the os prominens is definitive. It should be stressed, however, that in the accipitrid this bone is not of the usual sesamoid character. Rather it is elongated and acts in holding out the ertensor patagii tendon from the wrist joint. It has a distinct articulation with the end of the radius. When the wing is closed this ossicle is folded against the underside of the manus, and when the wing is opened it rotates outward and forward. A bone of the usual sesamoid type is described by Lucas (1882) for the great-horned owl. A "sesamoid" was observed by the writer in this species and also in the wing of Sagittarius; none was found in Faleo (as described by A1ix, 1874b). The typical oa pz,ominens is found only in the Accipitridae, although it varies considerably in slze; that of Pandion is little more than a sesamoid.

Tib iotarsus The distal end of the tibiotarsus is frequently fossilized. 2\ (224) M. JOLLIE

Accord.ing to Sushkin (1905:44-45) th9 ossified loops on the anterior alpect of the distal end can be use'd to separate the Accipitridae and Fal.conidae. 'In deveLopment,the loops are formed in connection with a soinelike ossiiication which ascends from the distal articular surface (figs . I?3, 179). This spine fuses with the lateral

FIBULA

INT.

Fig. L7g. Anterior aspect of distal end of left tibiotarsus of (right) and nestling (1eft) A. MiLuus migrane, B. FaLeo "aifttinnuniulis (after Sushkin, 1-905: figs. 15, 16) ' margin of the shaft and may isolate a lateral opening. An osseous 1;;; d".r"lops over the ertensor.dig?tonum eommuhis tendon to forn ihe^nedial bpening. The spine has-been homologized with the internedium (Morse, L872), 41n;ng tie catfrartids (f ig. 180) tlr" lateral opening is- absent; a remnant of it is indicated 6y a pit in GymnogVps. The tibio- does not tarsus of Neoeathartee (Wetmore,'The L944, Pl. V, fig. 1) rho* loop structure. tibiotarsus.of-Sagittarius has a small of""irrg" in^ sone specit"lt. The width of the osseus 19op is a i'"gg"iii"" featurê of this type. The accipitrids lack a lateral ;;;;i;;, afttrougfr fossa here appears in inany forms. The falconids " latirral gPelilg,- but Herpetotheree G;;iii'have fi"ll-developed -rn iftir niy be reduced" to a small hole ({VZ 85557). In Micrastur it is littie- more than a pit showing perforations ' Howard (L93ZzS7l bointed oul that, "The supratendinal bridge - is more horizontally'piaced in the Aegypiinae than in the eaglesr". to""""i, that of nebnbsyntee is directiy-comparable-loop with the nornal angte. The 6n1y really peculiar is that found in "iiipitiiacirahospiza aia Gymnogenys",in th'esil it is almost exactly horizontal i" potition. Theie tio lenera agree closely in thg forn of the iitiîot""trrt (as in other-1inb elenents), the bone having n-early p"i"f f"i sides. One specimen of teptodo.n was peculiar in having in inconplete loop--an anomalous condition. Othbr featurês of the distal end have been cited. Howard (l-932:57) noted that, "Th9 distal end of the tibiotarsus in the Aegypiinie differs from the Aquilinae and Buteoninae in greater FALCONIFORMES 25(225)

B l,ll A1}J # ffi K

#J J$ f N o

Fig. 180. Anterior aspect of distal end of right tibiotarsus of A. Sagittapius serpentarius, B. Cathartes auna, C. Gymnogyps eaLifornianus, D. Teratornis merniami, E. Leptodon paLLiatus, F. Chondnohienan uneinatus, G. Aeeipiter eoopenii, H. BusareLLus nigri,collis or ELanue Leueunus, I. AquiLa ehryeaetos, J. Neerosyrtes monaehus, K. Terathopi.us eeaudatus, L. Pandion haLiaetus M. Herpetotheres ca.ehi,nnane, N. Mi,erastuz' semitorquatus, O. PoLgbonws eheriuay or MiLoago ehimaehima, P. Mienohie?ar eaePules- eens, Q. FaLeo peregninua. depth relative to breadth. " A comparison of some ratios (Table 16) shows that the vultures do not differ materially fron other accipitrids. Deep articulations are usually typical of a cursorial type of leg such as is found in Sagittanius or a cathartid. The largest percentages among the accipitrids are found in some of the Aegypiinae and such genera as Accipiter, Circus, Haliaeetus, BueaneLLa, and Pandion. The distal articulations of Pandion are peculiar in every respect and certainly the leg is not cursorial. The falconids show a range of from 70 to 87% which places then in the same category as the accipitrid. Other characters as the length of the tibial crest are un- satisfactory critetia, as they cannot be accurately measured. 26(226) M. JOLLTH

Table L6. Percentage proportion of the depth of the distal end to the width of the distal end of the tibiotarsus. Distal end of Tibiotarsus F{ r{ (6 (d +J +) l, o ?n .F{ .Fl q) 'u E u (+{ || (d o o ,c P +J ! €rJ An, € .Fi tr oC lr Fo €q) I r{ F I t\ o EFA Cathartid Cathartes aura L3.7 LZ.8 93.s 12.0 1.1.6 96.7 Coragyps atratus 13. 3 L2.8 96.3 L2.6 1,3.1 L04.0 Sarcoramphus papa L9. 3 L7.7 92.0 Gymnogypscal i fornianus 24.3 22.3 95.7 Sagi ttar i id Sagittarius serpentarius 2L.7 2L.3 97.8 Accipitrid Pandion haliaetus L4.4 13.9 96.5 Pernis apivorus 11.0 7.8 70.8 Chondrohierax uncinatus L0. 5 7.2 68.6 Leptodon palliatus 8.6 5.7 66.3 Elanoides forficatus 9.0 5.8 64.4 Elanus leucurus 9.6 6.7 69.8 I ctinia misis ippiens is 6.8 4.4 64.7 Accipiter cooperii 7.4 5.8 78.4 Circus cyaneus 't '7 6.5 84.4 Haliaeetus leucocephal"us 15.6 rz.5 80.2 7L.6 Gypohierax angolensis 14.5 10.5 72.4 Gypa6tus barbatus 22.4 L7.5 78.2 79.0 Neophron perenopterus 20.0 L3.7 68.5 Torgos tracheliotus 24.5 19.5 79.7 2s .7 20.0 84. s Buteo jamaicensis L3.7 8.8 64.2 Busarellus nigricollis 12.7 9.6 75.6 Fa1conid Herpetotheres cachinnans 12.L 10.0 82.6 11. 5 10.0 87.0 Micrastur semitorquatus 11.8 9.6 81.4 Polyborus cheriway LZ.7 10,2 80.3 L2.3 L0.2 83.0 Falco peregrinus 11.0 8.1 73.6 Falco (5 species) 70.3-80.8, *************tt* ***** *rt****trtt***** Tibiotarsi are poor indicators of relationship and at best are usable only at the generic or specific level. Thus the distal end FALCONIFORMES 27(227 )

of this bone does not effectively separate the several types of falconiforms. The internediate condition' of Herpetotheres (note peculiar external condyle) between accipitrid and falconid appears to be a convergence. Two fossil cathartids , Phasmagapspatnitus and Palaeogyps prod,romus, were described by Wetmore (1927) on the basis of fron the Oligocene of Colorado. These tibiotarsi iiUiotarsi "P- pear to be cathartids U[t conceivably could belong to species il ieveral present orders, determination of their exact relationship being deilendent upon the other elements of'the skeleton. Cracraft and F.ictr-(L972) aSsociated Diatnoporni,s ellioti and Amphise-rpen' taniue eeilosseri with the cathartids on the basis of the distal end of the tibiotarsus (and the tarsonetatarsus in the case of the former). Their opinion may be justified, but I would prefer to know more elements of these species. Fibula The fibula is best developed among the kites and aegypiins; it nay extend nearly the length-of t!t9 tibiotarsus. The nost compllte exanple--there is considerable individual variation--is fouird in Pandion and Terathopius. In these genera the fibula has a distinct articulation with the tibial bone. In most genera' it ends well above the lateral condyle as a small lunp fgsed-with the tibia. The presence of the extbnded distal end in the kites and sone aegypiiirs is probably indicative of a more generalized state . Gymnognng",which resembles Genano.spizain many details, has the f6ngei, Leiter developed fibula of a kite. Foot The foot of the typical accipitrid is developed fol g-rasping, and this specialization-has resulted in strengthening of the claws and in the shortening of various segments of the toes. The extreme is achieved in nanpi-a" Stephanoa7tus, and Pitheeophaga' Most weak foot aicipiirias have sirong fe-etr ltt sone kites have a whicir resembles that of a chicken. Feet of a weak typer-perhaps useful in walking, are found in the cathartids (fig. 181-). On the struEiure of the foot the falconiforms can be divided into three groups (fig. 204): cathartid, sagittariid, and an accipitrid-falc6nid gioup. The first has long-front-toes' a very short hind toe, the lfawi of all digits are relatively shgr!..++ siraight, and ttte second and third-phalang-es of the.fourth digit group relatively are relatively long. The second- -has -short - toes, the hind tbe aplroximates the iecond gigi! in length, and the claws are fairly long and curved. In the first tyo gToupg ihe basaf phalanx oi the second 4igit is as- 1ong as the distal phalanx, and the basal phalanx of the fourth toe is as long or phalanx of that Tt-t" last- gToup ionget ihan the terrninal {igi!. - tenis to have longer, shaiper claws, the basq! phalanx of the second digit is dlstinctly shorter than the distal one, and the. Uaiaf phaianx of the fourth is distinctly shorter !1t." the terminal phalanir. The foot of Pandion is pe991i3r in that the outer toe ffVl is longer than the middle one (III) and is semi-reversible iit6 that of an owl. In addition all of the claws are equally 28(228) M. JOLLIE

\'-^

b=.n

of right foot of A . Sag.?ttariue Fig. L81. Bones the -eenpenta-1iu7 , B. Cathattes auna, C. generalized accipitrid (including Pandion)' D. Iehthyophaga iehthya'dtus, E. Aoieeda cueuloidEe ot ELanotdes forfieatis- or- Gymnogenys tApieus, F. generalized falconid. developed as contrasted to the preferential developnent of the hind and index claws in the other accipitrids. The aegypiin foot outwardly resenbles that of the cathartid in scalation lnd the blunt claws but has the typical accipitrid bone structure. Gypohie?an has well-developed claws, whereas Tenathopius resembles the aegypiin. The claws of the various accipitrids range fron the enormous curved'weapons of the large eagles through the short blunt talons of the aegypiins, to the slender, elongated tips of Pernis, Ggmnogenys- or the extremes , Rostrhamus and retinadtus. Of the feet examined only those of Miluus" fetinia, HaLtaeetue, and rcthyophaga showed fusion of the basal and distal phalanges of the second digit. Shufeldt (1891b) observed such fusion in specimens of fetinia and HaLiaeetus. The fusion in Haliaeetus may be incomplete with only the under surface involved, whereas FALCONIFORMES 29(229) the joint appears nore or less nornal above.

Sunnary and ConclusiPn The rniscellaneous skeletal parts have value when detailed agreenent is invoLved (at species or genus level) . , Lack of close slnilarity cannot be interpreted as more distant.relationship because of adaptive radiation. Any level of relation-ship_' or lack of it, is-possible under these circunstances. The three groups indicatel Uy the foot seem to have relevance since there Is ni real overlap'between then and rather basic agreenent within each style. OSTEOLOGY.. STIMII,IARY AND CONCLUSIONS

There has never been great doubt as to the intrinsic value of the skeleton in taxonomy. It has been used to support the association of different species as well as to demonstrate lack of interrelationship. It has generally played a supporting role to reLationships ba3ed on exteinal features. However' questions remain as to iis vaLue when species which are now considered as belonging to separate orders closeLy resemble each other in a partiEulir area or when species presume to belong. t-o the same -order differ narkedly. In the case of the falconiforns we are more involved with the latter than the former, but both are encountered and decisions must be reached regarding then. An exanple of the first situation is the similarity of.the cathartid to Cathanaeta, Maeroneetes or Fnegata. The association with Maeroneetes finds further support by the similarity of Teratornie to Di,omed,iain many features. ttri.s similaritl 9f form is not complete since nairy details, such as the ossification of the nasal passage area, the orbital gland inpress_ions, and the tarsOmetatarsus, c'an be used in their separation. However' we are comparing specialized end species rather than basic archetypes fbr ealh bf tne represent-ed orders. As represented in tfiA living fauna, the pr-ocellariiforn type is quite distinct fron the cathaitid, viewed in terms of the total mosaic of features, However, the of as is certainiy the pelecaniforn. -titilarity these terninal'generi with the cathartid cannot be dismissed since an exptanition based on adaptive convergence demands far too much. Somewhatless narked is the sinilarity of body and limb structure shared by the storks and Sagi,ttani'ue. Again there appears to be more than adaptive convergence involved. One can irrive at the assumption that there was 'some commonancestral type fron which thecathartid, sagitta-riid, procellariiforn, ("r4 p_erhap-solher groupl as f6iecaniform, and ciconiiform- - . -wel1) irose. From this stem type all of these highly sp-ecialized lines developed; yet each-retained a great deal of its common heritage. Jirst a! probable is the view that they underwent paral161 changes (bisically chargeable_to increase in size) so is to arrive it their present state of sharing many features in common. The nature of- the shared features suggests that they are a result of independent evolution rather than commonheri- 30(230) M. JOLLIE

tage, and this is further supported by their absences in the otherwise sinilar charadriimorph type (complex containing the Gruiformes and Charadriifornes of Wetmore, 1960) . Recognition of .functional nodification is difficult since so many facets must be exanined. Parallel changes might be charged to a number of general factors (increase in size), but demonstration of correlation would be difficultr,whereas more specific responses to needs, such as the relationship between a long neck and vulturine habits, are more apparent. 0n the basis of the nore obvious functional features, it night be assumed that the skull, the tarsometatarsus, and foot show the most adaptive modification whereas the body (shoulder girdle, sternum, and pelvis) is more conservative. The plasticity of the skull is explained through use of the bill as a feeding device. This in turn necessitates changes in the palate, prefrontal region, nasal passages, and olfactory area for movement or for bracing the bil1, modification of the eye and ear regions in response to the visual and auditory needs of securing food, and remolding of the cranium as a result of brain alterations, and changes in the neck and j aw musculature required for the new behavior patterns and movements. The weakness of this line of thought lies in the idea that the design of the body is not particularLy affected by changed feeding habits (but is affected by locomotor changes). It is certainly open to question which has been changed more in such groups as the anseriforn or the loons, the head or the body. It could be argued that once a particular specialized style of bill and feeding is achieved it is selectively "fixe-d" and undergoes only linited modification, while the body, which still serves a wide range of functions may speciaLize along any one of several lines. Here the question can be raised whether the bill of the stork could be modified into that of sagittarius in response to its predatory habit. This is improbable for two reasons. First, the stork bill should be viewed as a specialized end form (see below) and, second, sone of the storks are vulturine in habit and yet have lost none of their heritage. This type of bill, as well as the body form, is a dead end. It is unlikely that the foot of Sagi,ttaz,ius has been derived from that of the stork on sinilar grounds. It is more probable that a predator (bill and foot type fixed) developed long legs and a cursorial style of feeding. Whereas the bill of the stork is an extreme and distinctive structure that of the predator (1ike that of the charadriimorph complex) is of a supraordinal nature. The bill of the primitive avian had a hook at the'tip and needs only to be shortened and strengthened to be transformed (at least in its superficial outline) to that of a hawk or owl. Thus, in the case of predators one might assume that the body structure can be more distinctively affected by adaptive modification than the head. As can be observed, the bills of predators may differ at the finer levels of structure, but these differences are of the same order and kind as observed in the body. FALCONIFORMES 31(231)

When bill forms are comparable, the skull can be utilized for ordinal differentiation only by directing attention to what otherwise might be considered as rather minor details. For example, the skull of a bustard differs most narkedly from that of a charadriiforn in its holohinal nasal opening and in lacking supraorbital fossae for the nasal glands; the cathartid differs frorn the fulnar or albatross in its vestibular (alinasal) ossifications and lack of a bony voner. That such differences are occasionally uninportant does not mean that they are always so. In the situations cited it can be assumed that these details (correlated with other dissirnilarities throughout the skeleton) are of real value. In the.case of the parrot perhaps even such details have been effaced by modification of the bill; yet the fact that the parrot bill is only plastic within linits is attested to by the ease of itq identification and the wide range of food habits which it serves. A peculiar case, where the bill form rnight be considered comparable and yet there is almost total disagreement in its aspect, is that of the ibis and stork. These two essentially agiee only in being "desmognathous." A, nuch better cornparison of Uiff form exists between the ibis and the scolopacid than between the ibis and stork. The major disagreement is that the scolopacid the palate is schizognathous. I doubt whether the condition of - is worthy of recognition in this particular case, the scolopacid being more like the ancestral source of the ciconiiforns in terns of bill structure. A consideration of the variations of the skull, which should be plastic to adaptive modification, tells 1ittle. The bill' as the nost responsive part, te1ls least. Yet i,n spilq of this, the skull suggests that there are four types of falconiforns, and these show 1ittle or no "intergradatj-on" or phylogenetic interrelationship. If these four types of crania could be adequately characterizbd the continuing discussion of the lnakeup of this order would cease, but they cannot. Thus the answer to the riddle nust be sought elsewhere. The mandible of the bird has never been thoroughly studied since, in its fused adult forn, few useable features can be detected. The mandible of the falconiform gives hints but nothing mole. One is cautioned by the possible parallelism in form of that of the Sphenisciformes, Gaviiformes, Procellariifornes, Pelecaniformes, and Ciconiiformes The tarsometatarsus, like the skull, should be quite responsive to adaptive changes in the foot dictated_by-its use, but the nunber of conparable points is restricted. The four styles of the falconiforms support this view, and the fact that any one of these is not well natched in any other order gives added significance. In contrast some types of tarsometatarsi (those of the charadriinorphs) can only be considered as supraordinal. The falconiform tarsometatarsi share only one feature: the niddle trochlea does not project as strongly as it does in most birds. This sirnilarity is of little inport, since there is lack of agreement in detail, and the general type is rnatched in 32(232) M. JOttIE some non-falconiform groups; it might have functional meaning rather than close relationship. There is some agreement in that the hypotarsal structure is seldom perforated by tendinal canals as it is in many other kinds of birds. There does not appear to be any funct i.onal corrclation; this lack of canals appears to be a primitive feature shared with mtrny other kinds of birds. Other elements of the skeleton, being simpler, ol more constant, in form are less useful than the skull or tarsometatarsus. The sternum, which is frequently ignored, seems to be of value in the identification of a columbiforrn-galliform complex. (The sinilarity of the pigeon and the chicken does not appear to be convergence, although it could be interpreted as such.) In contrast, the sternum does not support the union of the gruiforns and charadriiforms; in the former it is a long narrow structure, whereas in the latter it is short and broad. The elongated forn is probably a specialization achieved by the ancestor of both the cranes and rails. Internediate forms do not connect these two types; rather other species with elongated sterna appear to re- pii:sent modifications- in the same direction as that taken by the ancestral gruid (i.e. parallelisns). Whereas the sterna of some groups (orders) are strongly differentiated those of others are not. Within an order, the passeriforms or cuculiforms for example, two styles.of_posterior margins occur with some internediate species or individuals. Among the ciconiiforns one finds both styles but no identifiable inter- mediates. Arnongthe charadriiforrns there is general agreemelt with some quite distinctive variants (PhiLoheLa, Seolopar, Burhinidae, Parridae, Rostratulidae, Alcidae, Thinicoridae) . The pelvis has a f.imited number of conparable features, which appear to-be quite plastic. Particularly interesting are the close agieenent of pelves of diving types. Although the pelvis has little vilue anong the existing birds, it appears to support the early dicotomy of birds into the Palaeognathae _and Neognathae. Sulnrnarizing the case of the falconiforms, the skull is somewhat suspect as to value, particular1-y the forn of the rostrum. In contrast, the sternum and shoulder region more accurately reflect the heritage of the four types described. The tarsonetatarsus and sone of-the features of the skull support the evidence provided- by the body. The cathartid resembLes some species of the procellariiforms, -enoughpelecaniforms, and ciconiiforns. The resemblance is not exact to demandunion into a single group; each is equally-dll- tincl from the others. These groups are each composedof highly specialized species in terms of size and habit, and the entire aisemblage shows many commonfeatures rvith a less specialized type, *hich can be represented by Catharctcta ekua, a memberof th;, charadriiform complex. The closer sinilarity between the extrene species (Teratornis to Diomedna) in each of these groups is as much parallelisn as anything. There can be little doubt that each of these orders represents a prinary line of radiation from the ancestral bird. Conparing the cathartid with the other falconiforms produces no evidence of any closer tie than being b irds . FALCONIFORMES 33(233)

The sagittariid typer or the basis of its body skeleton, appears to be like the storks. There is 'some agreement with Cardama, but this i's.more likely convergence. If one disregards the presuned convergent features there is Little evidence of stork affinity. One can assume that this aberrant genus does, in fact, represent a product of a pre-accipitrid line, but such a speculation finds no osteological support. To argue that this type comes closer to the accipitrids than any other and, therefore, that this inplied similarity is evidence of relationship, disregards the posibility of convergence. The question then follows whether the Long legs of Sagittariue preceded the hawk-like bill, whether the reverse is truer or whether these features were developed simultaneously. I have concluded that the bill and feet, as paralle1 developnents to the accipitrid, preceded the long legs. The accipitrid is one of the most distinctive types among birds, Iet it has aLways been linked with the fal-conid. Its individuality suggests that, like the sagittariid, it is an independent l"ine-of development fron a very basic stock. Simi- larities in shoulder girdle to the storks (or cathartids and sagittariid) nean f.ittle when viewed in respect to the total vaiiation in these structures and are as attributable to functionaL value as to comnon ancestry. Pandion appears to be an aberrant accipitrid and as such is usefuL in testiirg criteria. It differs consistently fron the of these typical accipitrid-is in every feature, bt! the nature di?ferences such that interrelationship can be assumed. Con- pared with the cathartid or sagittariid it is qppalent that a iifferent degree of differentiation is involved. For example, the accipitrid tirsometatarsus is defined by the separation of the hypotarsus into nedial and lateral calcaneal processes. Pandion cblnes as close to this as some of the kites, but the other types do not even approxinate it. Ienatornii menriami, has a sinilar relationship to the cathartids, sincer irs a highly specialized specigt, it is not so different fron the typical cathartid as to denand separate consideration. It i's- tite conparing a pelican to other pelecaniforms or an albatross to other procellariiforns. Using the aberrant-genus nethod of neasure, each of the types described takes on greater significance. The falconid type is most distinctive on the basis of its shoulder girdle. The sternum and tarsometatarsus are rnarkedly different. Their closest approach is to the owls, oI parrots, but even here the Leve1 of lgreement is low. Herpetotheres is the mo'st accipitrid-1ike, but it appears to represent convergence rather than an internediate forn. Its affinities are entirely falconid; in no detail is it truly accipitrid. Although the falconid might be assumed to be less specialized than the accipitrid, theie is no agreenent with the sagittariid or cathartid.

I{YOLOGY

Introduc t ion 34(234) M. JOLLIE

There is need for a revised systen of nanes for the muscles of birds. The general treatises, those by Stresemann {L927-t934), Grass6 (L950), and George and Berger (196'6), do not satisfy this need because like the older account of Gadowand Selenka (1891), they are not in harnony with my view that a system more nearly like that of the mammalis desirabl.e. I believe that a basic systen of muscles, developed in the origin of the amphibia, has been modified in each of the derived classes (Jol1ie, L96Z) and that the individual muscles can be identified with N.A. (Nomina Anatomica) terms. The utilization of names conpounded from origins and insertions, as advocated by comparative anatomists of ttre last century, results in a different system of names for each class and makes conparative study difficult, if not inpossible. The assumed impropriety of comparing muscl-es of two classes has been based in part on innervation. It is now apparent that the nerve supply nay be no more reliable than the nuscLes thenselves. Since holnology cannot be irrefutably demonstrated it should be sufficient to look for a less exacting correspondence of parts, in agreement with their evolution fron a common(or nearly conrnon) ancestral pattern. There is Little question that "homologous musclesf in different vertebrates vary in their interrelationship from exact correspondence to rough resemblance or no resemblance at all. The use of a systen such as I advocate does not inply exact correspondence- of parts, but it does imply sone level of corresiondence in agreelnent with the taxonomic separation of the groups. Such a system does not compronise exact scientific procedure, but rather it facilitates by increasing llle ease of comprehension and connunication. By abandoning rig_id and meaningless concepts of homologY, less tine is needed for argument on-terminology and nore tine is available for the exploration of variation within groups. Certainly little is to be gained by detailed conparison ol the muscles of different classes with the goal of demonstrating their dissinilarities instead of their evolutionary continuity and adaptive specialization Studies of the myology of birds are few in nunber. Several of the older treatises are those of Owen (1849) fot Aptenya, Macalister (1864-1866) on the ostrich, Alix (187aa) on a timamou, Perrin (L875) on the hoatzin, Watson (1883) on pengui-ns, Shufeldt (1890) on the raven, Marshall (1905) on the poorwill, Chamberlain (1943) on the chicken, and Fisher and Goodrnan(1955) on the whooping crane. Summaries of the nyology of birds have been presented-by owen (L866),-Stresenairn by Alix (1874b), Gadowand Selenka (1891), Beddard ifAgS), (L927-1934), Grass6 (1950) and George and Berger (L966). Several authors have dealt with restricted portions of the muscular systeln and these include Garrod (1873-L874; thigh), Fiirbringer (1-888, L902; shoulder and breast), Buri (1900; wing), Lakjer (L926; jaw muscles), Banzhaf (1929-1930; wing), Hudson (L937, L948; Hudson and Lanzilloti 1-955; thigh and wing) , Fisher (1946; wing and leg) ,. Beecher (1950,- l-953; jaw-muscles) and vanden Berge (1970; appendicular muscles of ciconiiforms). FALCONIFORMES 35(235)

Materials The honologies (often tentative) irnplied by the names used in this account were deternined by dissection of more than 50 genera of birds, three of reptiles, and three of nammals plus a review of a part of the f.iterature on the nyology of these classes. The ilLustrations will assist the reader in identifying the muscles described with the nanes used by other writers. Conparisons within the faLconiforms are based on 24 genera dissected as indicated in Table L7. The source of supplementary infornation is also indicated in that table. TabLe 18 lists the nuscles of the bird and the assumed corresponding nuscle of the hunan when one is present. Sone of the nuscles of the human are not found in the bird and this also is indicated in the table. OnLy those muscles starred are discussed in the following section; the others are not as yet known to show useable variations, although some of them nay. * * * * * * ** * * * * * * * * * * ** * * * * * * * * * tr t! tt * Table L7. Falconiforms dissected for this study or described in the literature.

Symbols: A - Alix, 1874b; B - Beddard, 1898 and 1903; b - Berger, 1956b; F - Fisher, L946; f - Fiirbringer, L888; G - Garrod, L873, 1874; H - Hudson, 1948; M - Mil.ne-Edwards, l-867-I87L; N - Nitzsch, L863, 1866. The species personally investigated are indicated with a J. Colunn I II III IV V VI a ox dbo tr d uo o o .Fl -o U) F{ .r.l F 5 O b0 b0 .'-{ E\ o o F{ Ehh t{ r{ O'-{ h F{ O +)(tl O c0 .r.,1 O*r € .r{ f{ x EO F{ b0 O .( o+r a d {J oo UO+JA.r{ h qt{ t tn o .q h Fl O A A A k U) Cathartid Cathartes aura JF f H BJ Coragyps atratus JF F H G J Sarcoranphus papa F G B Gymnogypscal ifornianus F Vultur gryphus F Sagittari id Sagittarius serpentarius HB G Accipitrid Pandion haliaetus J f H J Elanus caeruleus J J Ictinea plumbea J J Aviceda (cuculoides) B Aviceda subcristata J J Chondrohierax uncinatus J J Gymnogenystypicus B B B Milvus milvus G 36(236) M. JOLTIE

I II III IV V VI Milvus migrans J B Haliaeetus vocifer' G Haliaeetus albicilla f B G B Haliaeetus leucocePhaLus J J Gypohierax angoLensis B B Gyia6tus barbatus N B B Neophron perenoPterus - Dryotriorchis sPectabilis B CircaEtus (gallicus) B Haenatornis cheela G TerathoPius ecaudatus B G B Aegypius monachus B B foi-girs trachel iotus J B G BJ Gyps fulvus N B G B Necrosyrtes monachus J J Morphnus guianensis B Harpia harPYja B G B Harpyhaliaetus coronatus B ncciliter striatus J Accipiter nisus Accipiter fasciatus Acciiiter cooPerii J H Accipiter gentiLis G Melierax musicus B BJ Circus cyaneus J H Circus aeruginosus G? B Circus Pygargus G Circus maurus B Kaupifalco monogrammicusJ G J Hvpbnorphnus urubitinga B (albicollis) B t iircoptbrnis B Buteo fuscescens Buteo buteo B G B Buteo jamaicensis J H J Buteo swainsoni H Buteo lagoPus G B Buteo nitida B fasciatus B Hieraeetus B SpizaEtus sP? Lopha6tus occiPitalis Aquila raPax Aquila heliaca Aquila chrysaEtos J H BJ A{uila audax J M GJ Fal conid Herpe to theres cachinnans J BJ Micrastur ruficollis J J Milvago chimachima J B BJ chimango B B Milvago J Daptrius ater Polyborus Pl ancus Polyborus cheriwaY BJ FALCONIFORMES 37(237 )

I II III IV V VI Polihierax senitorquatus J b J Microhierax caerulescens B Falco peregrinus tl G B Falco rusticolus B Falco columbarius B H B Falco biarmicus G Falco jugger Falco mexicanus Falco subbuteo G Falco sparverius J H J Falco tinnunculus A B G B Falco vespertinus B Ieracidea berigora BJ ************* * * * * i * * * * * * * * * * * rt tr tr Table 18. The honoLo9ies of bird and hunan rnuscles. Those narked with an * are described in the text.

Bird Human Skeletal ltueelee of Head Muscles of Eye rectus inferior rectus inferior rectus nedialis rectus medialis rectus lateralis rectus lateralis obliquus superior obliquus superior obliquus inferior obliquus inferior orbicularis oculi orbicularis oculi rectus superior rectus superior quadratus nictitantis Lacking pyranidalis nictitantis lacking levator palpebrae levator palpebrae superioris depressor palpebrae lacking pars ventralis pars posterior Jaw and Throat Muscles *nasseter et temporalis masseter, temporalis pars temporalis et quadratis pars orbitomandibularis posterior pars orbitoquadratis pars quadratomandibularis pars orbitonandibularis anterior pterygoideus pterygoideus exter- pars'dorsalis nus, pterygoideus pars internedia internus, tensor pars ventralis tynpani pars pterygoideus *depressor mandibulae ?I acking ceratornandibularis ?1 acking ?lacking diagas tri cus 38(238) M. JOLLIE

*ceratoglossus genioglossus, hyoglos- ' sus, chondroglossus, s tyloglos sus , glossopalatinus depressor glossus levator glossus thyreohyoideus thyreohyoideus thyroarytaeno ideus thyroarytaeno ideus constrictor glottis arytaenoideus obliquua, transversus, aryepiglotticus cricothyreoideus cr icoarytaeno ideus posterior, cri coarytaeno ideus lateral i s, cricothyreoideus geniohyoideus geniohyoideus tensor columellae stapidial muscle Denmale and Conetri etor of Head lacking + fauces muscles, I pharynx muscles, I facial nuscles ceratohyoideus (?) s tylohyoideus *transversus hyoideus 1ack ing intermandibulari s mylohyoideus *constrictor coll.i platysma Denmale of Body *sternocle idoccipital is sternocle idomastoideus, pars ternporalis omohyoideus pars clavicularis pars trapezius pars propatagialis pars spinalis *derrnocervical is lacking *sternohyoideus *s ternotracheal is f sternohyoideus tracheal is f sternothyroideus *tracheobronchial is pars ventralis brevis pars ventralis longus pars ventrilateralis pars dorsalis longus pars dorsalis brevis pars' dorsilateralis *cutaneous maxinus (lacking cutaneous pars latissimus maximus of other pars metapatagialis narunals ) pars axillaris pars abdominalis *expansor secundariorun 1acking adductor rectricun lacking cloacal aperture sphincter lacking FALCONIFORMES 39(239)

lacking corugator cutis ani, sphincter ani externus, sphincter ani internus rtransversus perinei I ack ing Skeletal Mudelee of Ventebnal Column longissimus dorsi longissimus dorsi longissimus cervicis et capitis longissinus cervicis, longissinus capitis spinalis dorsi spinalis dorsi spinaLis cervicis et capitis spinalis cervicis, spinal-is pars cervicis et capitis capitis, (splenius capitis) pars spinalis cervicis pars cervicalis anterior pars cervicaLis posterior rectus capitis posterior najor rectus capitis posterior naj or rectus capitis posterior ninor rectus capitis posterior minor lacking obliquus capitis inferior, obliquus capitis superior nultifidus et senispinalis dorsi nuLtifidus, senispinalis dors i semispinalis cervicis 1 ( senispinalis cervicis, rota- t 1 tores, iliocostalis obliquotransversales J \ cervicis, splenius cervicis senispinalis capitis senispinalis capitis il"iocostaL is lunborun iliocostalis lumborun, iliocostalis dorsalis longus lateralis t intertransvers ares intertransversares ) Longus colLi longus colli pars thoracicus pars cervicalis longus capitis et rectus capi ti s longus capitis, rectus anterior capitis anterior rectus capitis lateraLis rectus capitis lateralis interspinales interspinales SkeLetal IaiL Mueelee levator coccygis lateralis caudae dorssalis a1i s f coccygeus *depressor caudae l 1 levator ani pars superficialis I div. levator cloacaeacae I div. retractor cloacaeloacae pars. profundus I lateralis caudaecaudae ventralistralis I pars depressor coccygiscysls pars infracoccygis J SkeLetaL Body llalL Mueelee quadratus lumborum quadratus lumborum scaleni et levatores costarum scalenus anterior, scalenus nedius, scalenus posterior, levator costarum 40(240) M. JOLLIE

intercostales externi intercostales externi intercostales interni intercostales interni, pars vertebraLis subcostales pars sternalis pars internus *s ternocoraco ideus lacking . pars sternalis pars costalis costisternalis lacking lacking serratus posterior superior, serratus posterior inferior obliquus externus abdoninis obLiquus externus abdoninis obliquus internus abdominis obliquus internus abdoninis, cremaster rectus abdoninis rectus abdominis, py- ramidalis transversus abdoninis transversus abdoninis costopulmonari i (? diaphragm) SkeLetal SuPenfi,eial Mueelee of Peetoral Limb rhonboideus rhonboideus najor, pars superficialis rhonboideus minor pars profundus serratus anterior serratus anterior pars superficialis anterior pars superficiaLis Posterior pars profundus I acking levator scapulae *latissimus dorsi latissimus dorsi pars anterior pars posterior Skeletal DoreaL Mueeles of Uppen Arm *teres major teres najor *subcoracoscapul aris subscapularis, (sub- pars subcoracoscapularis c1avius ) pars scapularis anterior pars scapularis posterior *deltoideus deltoideus pars clavicularis insertion longus, insertion brevis pars scapularis *teres minor teres minor pars externus pars' internus *triceps brachii triceps brachii pars externus pars internus pars axillaris anconaeus anconaeus FALCONIF'ORMES 41(241)

Skeletal Yentnal Mueelea of ttppen Arm *pec toral is pectoralis major, pars superficialis pectoralis ninor pars profundus pars patagialis *supracoraco ideus supraspinatus, infraspinatus coracobrachialis brevis'l coracobrachial is *coracobrachialis longus / *biceps brachii biceps brachii pars biceps brachii pars propatagialis brachial is brachial i s SkeLetal Doreal Mueelee of Foneanm lacking brachioradial is extensor carpi radialis brevis extensor carpi radialis brevis extensor carpi radialis longus extensor carpi radialis longus supinator supinator extensor digitorum conmunis extensor digitorum com- pars anterior munis, extensor digiti pars posterior quinti proprius extensor indicis extensor indicis proprius f abductor pollicis longus, abductor et extensor pollicis 1 extensor pollicis longus, I extensor pollicis I brevis extensor carpi ulnaris extensor carpi ulnaris Skeletal Ventral MuseLes of Fonearm pronator teres pronator teres fLexor carpi radialis flexor carpi radialis flexor digitorum superficialis flexor digitorun sublimis, palmaris longus flexor digitorum profundus flexor pollicis longus, pars anterior flexor digitorum pars posterior profundus pars ulnaris flexor carpi ulnaris flexor carpi ulnaris pars anterior pars posterior lacking pronator quadratis Skeletal Donsal Museles of Manus interossei dorsales interossei dorsales pars abductor digiti II pars extensor digiti II pars extensor metacarpi pars interosseus dorsalis pars flexor metacarpi pars flexor digiti IV, flexor brevis digiti IV 42(242) M. JOILIE

abductor policis brevis opponens pollicis flexor pollicis brevis adductor pollicis palnaris brevis lacking abductor digiti quinti flexor digiti quinti brevis opponens digiti quinti lunbricales interossei volares (part of) interossei pars interosseus pal"naris volares SkeLetal Ventral Mueeles of Manue metacarpal flexores pars flexor digiti I I \ pars adductor digiti II I lacking pars abductor digiti III I pars abductor brevis digiti III ) Skeletal Doreal Mus eLes of nhigh sartorius s artorius *tensor fascia latae tensor fascia latae i 1io fibularis (short head of) biceps femoris proxinal deep layer *gluteus maximus gluteus maximus gluteus nedius gluteus medius gluteus rnininus gluteus minimus *pirifornis piri forni s i1 iopsoas psoas major, psoas minor, iliacus. lacking pectineus distal deep layer *rectus femoris rectus femoris *vastus lateralis vastus lateralis *vas tu.s intermedius vastus internedius, articularis genu *vastus medius vastus medius Skeletal VentraL Museles of Thigh *biceps fenoris et seninembranosus pars lateralis (bicePs femoris) (long head of) bicePs pars rnedialis (seninenbranosus) femoris, senimembranosus *semitendinosus semitendinosus *gracilis et. adductor longus pars gracllls gracil is pars. adductor longus adductor longus *adductor brevis et magnus adductor brevis, adductor magnus, (adductor ninimus ) obturator internus obturator internus lacking genellus superior, gemellus inferior, quadratus femoris. FALCONIFORMES 43(243)

obturator externus obturator externus pars internus pars externus *caudil iofenoral is I acking pars caudatus pars iliacus SkeLetaL Fleaor I,Iueeles of Shank *gastrocnemius gastrocnemius pars lateralis pars medius *plantaris plantari s pars accessorl.us pars tibialis *soleus soleus *fLexor digitorum superficialis lacking pars perforans et Perforatus digiti I I pars perforans et Perforatus digiti II I pars perforatus digiti IV pars perforatus digiti III pars perforatus digiti II popl i teus pop1 i teus *flexor hallucis longus flexor hallucis longus *flexor digitorum longus flexor digitorun longus lacking tibialis posterior Skeletal entenso? Museles of Shank *tibialis anterior tibialis anterior *peronaeus longus peronaeus longus lacking extensor hallucis longus peronaeus brevis peronaeus brevis *extensor digitorum longus extensor digitorum longus peronaeus tertius Skeletal Tarsal Muselee *extensor digitorum brevis extensor digitorum pars hallucis brevis pars abductor digiti I I pars extensor digi'ti I I I pars adductor-extensor digiti IV abductor hallucis, flexor digitorum brevis, abductor digiti quinti, quadratis plantae, ad- 1acki ng ductor hallucis, flexor digiti quinti brevis, interossei dorsales *interossei plantaris interossei plantaris pars' abductor digiti IV (pars adductor-flexor digiti IV) pars adductor digiti II *flexor hallucis brevis flexor hallucis brevis pars externus pars internus *lumbr icales lunbricales 44(244) M. JOLLIE

Comparative Anatomy Masseten et Tenponalie (adduetor nandibulae) The jaw muscles of birds have been studied in detail by Lakjer (L926), Starck (1940), Beecher (1951, 1953), Barnikol (1953b) and Starck and Barnikol (1954). These studies have not convinced me of the usefulness of Lakjerrs terrninology or the taxonomic value of the,subdivisions of the nain adductor. The parts identified here are equivalent of Lakjerrs (1926) names as follows z pa?s temponali,a et quadratie = adduetor mandi,bulae (eaternue and p.osterio?) , pa?s orbitonandi.bularis .= pseudotem- itonalis eupenfieiaLie, pa"s orbitoquadratis = protraeton quadrati, pqrs quadratomandibuLaria = paeudotemporaLis profundua, Starck and Barnikol (L954) studied the jaw rnusculature in (aura?) Sareoramphue Aceipitet genti.Li.a Aeeipiter Cathaptee , PaPa' " nieua, Buteo buteo, Haliaeetus Leueoeephalue, Iorgoe traeheli'otus" GApe fuluue, and Faleo tinnuneulus. There is fundanental ,agree- _ m6irt imong lfr of the genera exarnined by these writers and myself, the nain differences being in proportions. For exanple the pars orbitomandibularis of Elanus is small and tapers to a thin insertion tendon whereas in other falconiforms this nuscle is bulky. rt mavor nav not tape'rt3r2"!3|uio2;oubulae In Cathartee and ConZgyps, the origin of the depnesso? mandibulae extends over the ventral portion of the preceding muscle (figs. 68, 76, 94r 114). Cireus shows a separation of this muscle ind-the tbnponaLis, which are in contact in most sPecies. This separation ippears to be the result of the enlarged and protruding tyrnpanic nargin. CenatogLo s sus The position of the eenatoglossus (fig, L82) and its relationship with the translersus hyoideus varies as indicated under that musLl-e. In the cathartid, sagittariid, and sorneaccipitrids (Elanus" Pandion), it arises along the lateral and ventral nargins

D INTERMANDIBULARIS

CERATOMANDIBULAR RATOGLOSSUS

TRANSVERSUS HYOIDEUS CERATOHYAL

CONSTRICTOR STERNOHYOID COLLI B

Fig. L82. Ventral view of hyoid-nusculature of A. generalized acEipitrid and falconid, B. cathartid, C. Pandion haliaetue' aui,cbda auberietata o.r Neeroeyntee monaehue. is best developed in the cathartid, the vultu_rine species of the a-cipitrid, attb prLyborus as a corollary to the extensibility of the throat'which noimal.ly is more pouch-like than in the other falconiforns. However, it is equally well developed in H^enpe- and Uicraetur'. In Coragype and Uiluago some of its toiheree ' f ibers (platyema myoi.des of Or*en,-1gOO)attached to the fascia lying md'diat-to th; hyoid cornus . This pa{t of the muscle, and pirtlcularly its postbrior part,- is- least deveLoped in FaLeo. it is continuous posteriorly with the rest of the constri,etor (or ephineter)-E"-.The eolli' of the neck. pare eolli is best developed in the cathartid (fig. 183)' and somewhat less developed in fon:gos or Neetosylres-. It is just distinctive in HenpetothZree and Aie?astVo, P?rtlcularly behind the heaa. tnis layer is present in aLL, bt-t-sometimes it is hard to observe being 6ut one- layer of fibers thick and with the fi'bers sometimes widely spaced and adhering closely to the skin. tiz":12'"X311t;'r3'"I7lir"ro" rn thecathartid, (rig. r.84)is best developed as a continuous layer serving the entire area of the neck. This layer of muscle is well developed in all of the falconiforms and in all the pa?s tenporali.s and pans eLatsieuLaris are continuous through a part of their fibers. The pars trapezius is present in all and a true trapezius appears to be present in 46(245) M. JOLLIE

IOHYOID TOHYAL

STERNOHYOIDEUS

Fig. 183. Ventrolaternal view of throat musculature of Cathartes au?a.

some. It is weakest anong the accipitrids where only a few fibers atiach to the head of the clavicle. The pars p"opatagiaLis is o"iy weakly developed (the skin of the shoulder region st-rongly bouira to the underiying muscles) and difficult to identify in itrii-gio"p (observed i; Cireua, Falco,.and fenaeidea). In none of these do ttr,i muscle fibers extend to the tenaon patagii, tendon. Tie pare spinalie is not present in any of the falconiforms. DermoeenoieaLie The dermoeenuiealie is best developed irt Neenosytnes (fig' 184) where a series of bellies arise frbn the fifth to the ninth or the tenth cervicals. This muscle is present in all falconiforms has only one side. but usually "trr2i.r"71:lto"::" The e ternohyoid.eua is'best .Jeveloped in the cathartid (figs. large-in the vulturine accipitrids, S-agittartus" L83, 184), -and nuii"tothLres Ai.eraotun, and least.developed in lh" tlPical accipitrids and falconids. In Cathartes and most falconiforms,, partially that of sternoeLeidoeeipi-taLis; its brigin underlies -the such is. not the case in CoragAps. The belly fans ou! over the skin of the neck then tapers-ib the insertion. In the cathartid, the insertion is along the hyoid cornua with occasionally 1n . of tlu bundle attiching at !h" posterolateral nargin "tirt"tticthyroid cartilage. In Sagittani.us, there is a lateral insertion ;;'ih;-hyoid coinus and a narrower nedial one on the Posterolateral *-rgin of the thyroid cartilage. The accipitrid- and falconid agree in fiaving a disp'ersed insertion along lhe lateral nargin.of the thyroid lartila'ge which extends forward to the basihyal in common FALCONIFORMES 47(2\7 )

STERNOCLE IDOCC lPITALIS PARSTEMPORALI

DERMOCERVICALIS

PARSTRAPEZIUS

TERNOHYOIDEUS

PARSPROPATAGIALIS

PARSCLAV]CULARIS

Fig. 184. Neck dermals of faLconiforms, based prinarily gt Ca-thartes aura which actually lacks the pare p?opatagiaLi,e shown here. with the thyreohYoideus. Sterno traeheaLi s In the cathartid the sternotracheali,e arises on either side of the nidline from the dorsal nargin of the sternum. The two bands are in contact nedially and pass forward and dorsally to the ventral aspect of the trachea (still in contact, fig. 200). They blend with the tnaehealie of one side. In the accipitrid the'origin is somewhatmore lateral from the connective tissue overlying the sternocoracoid muscle (indirectly from the margin of the sternum, fig. 189). In Pandion these muscles insert upon the trachea asynmelrically but in contact one with the other. The muscles of the falconid or sagittariid arise laterally fron the anterior margin of the tip of the sternocoracoid process of the sternum (as in the najority of birds) rather than fron the sternal. margin behind or medial to the coracoid. The bands in- 48(248) M. JOTLIE sert laterally on the trachea and blend with the tnaeheali.e. Traeheobronehi.aLi-e Among the falconiforms, the tnacheobronchialzls.does not show a y subdiiision but there are variations in the,syri,lge?l-1"- i"itiot which have been described by Beddard (1898:475-476) . The cathartids differ from the others in lacking a distinct syrinx and tpaeheobronehi.aLi's-Th; muscle (fig.-is-more 200) ' iyiitr* of Sagittani,us boxLike than that of the accipiirid'and the nuicLe inserts upol the second bronchial serniiing. In-th; accipitrids the oval gap,between the second and ifrira sErnirings, occupied by the interannular rnembrane,is much ieaucea and t[e'nusc16 inserts on both of these sernirings or on only the second. In the falconid the interannular menbrane is large. Tltu traeheobronehialie of Faleo i.nserts on the interannular membrane (on a transversely el"ongated cartilaginous bar, no! observed in i,aleo sparuerirzs)'betweln bronchial semirings 2 ald_3 (eee Sushkin 190Sr6i-,' fig. 23). The various spe-ies of FaLeo are alike, fJt the bar, and igree with Polyborua" Daptnius, Miloago' riraZidea,"i."iii and.polihierarl fn Herpeto.-bhereethe insertion is on ih; seiond bronchial semiring, and-s1ightly oq the first, whereas Aierastur is intermediate wilh a part of the insertion on the interannular membrane cutaneou. Maxt)mus The pans Lati,ssinue of the eutaneous masimue-is present in nost fafcbniforms--vestigial in Hetpetothenes and Aieraatur poli.hi.erai 1956b:327). It i;;i .Ut""t in as suggesied bI ?u1ggr, riay be pres"tt ot absent (extremely vestigial?) in what may be ioitr ideied related genera- - Coragype-Cathantes, Neenoeynt"g-Tongos' ui'i"iiv"tut is pecuiiar in that this muscle arises over the an- dorsi (also ierior"part oi itt" posterior slip of !f9 latissinus the case in Sigiitai'tue). cymnogyp.e-differs in having- this muscle divided into tiuo p"tts inseriing-i'ith-the two parts of the pars. iutii"i"gi,aLi.i', Tigtttarius differs in the tr-ansverse orientation of ihe Ulrry (usualiy directed anterolaterally)' pans'i"iopoiagialie arises by to four narrow heads The lwo (in from tibr'three and folrr or fron each-of the first four ribs FaLeo or Gampaonyoan aponeurosis adds ribs five and six) ' Thq otiLi"r L;l;il' those'of the serratus an-teri.or pars- superfieiali,s. nerfietotheres."iu is peculiar in the-origin of this muscle fron two ;it; and in its insertion extending iron the underside of the scapular tract of feathers almost to the nidline (insertion neaily matched in GanPsonYn). in. ihe poo" aoillar|e and'pars abdomi-nalie are apparent only - the faicohia (iig. iASl. In Faleo,or Ieracidea, the.pans ariLLaris arises from tlie Suter 6nd of the deltoid crest, at the outer of the the axillary-connective *.tgi" of tn"--inrJrtion -peetoralisi tissue brace i;-;;it riigntfy dlveloped. In the other falconids, ift"-oiigi" is sornewiratm6re iosterioi from the connective tissue and fascia of itt. pietoraLds^margil between the peetoralie and t;;"p;-.- The axilLiry connective-tissue brace is stronger with !h" posteriorly to brace and result that the belly extends first -the then ventraLfi to it; insertion on the underside of the posterior FALCONIFORMES It9Q49)

Fig. 185. Ventral view of body of Faleo eparoerius showing pars asf,Llarie and pans abdoni.naLie of the eutaneoue masi.mus. end of the axillary division of the ventral tract of feathers. Reduction of this part is greatest in Herpetotheree and Mi,erastur where only the posteriorly directed part of the belly is present-- no ventrally directed fibers.were observed in the connective tissue brace. The pars abdoninalie is weakly developed in all falconids; it is vestigial in Herpetotheres and t'tieraatur. In the other groups, the skin is usually closely bound to the peetoraZie rnuscle in this region thus elininating the apparent function of these nuscles. A vestigial pa?s asillarie was observed in Kaupifaleo but in no other accipitrid. Eapans or Se eundario rum The espansor eeeundariorum (see discussion in Berger, 1956a, and Jollie, 1957b) is well developed in Sagittariue and the falconid (fig. 188). In these a connective tissue brace (tendon) is associated with the nuscle and extends from the axilla nearly to the el.bow. Most of the muscle fibers arise fron this brace whereas others rnay arise (falconid) from the skin nearby. These fibers insert along the outer three tertiaries and the inner secondaries. Other "bel1ies" of this conplex serve the other tertiaries. In the falconid, the bracing band (tendon) is an- chored in the axilla by the general connective tissue around the nerves.and blood vessels. Usually at leasttwo bands converge here to form the brace. The main one passes through the group of blood vessels and nerves and anchors broadly over the inner aspect of the eubeoracoaeapularie to the angle of the coracoid and scapula. The other extends inward along the path of the blood vessels and nerves and disappears on their surfaces. A third band from the nargin of the head of the biceps is usually identifiable. In Sagittariue the main attachment is to the posterior nargin of the subcoraeoaeapulanie pa?s ecapularis poaterior. The cathartid is much like the falconid but the belly and its 50(250) M. JOLLTE bracing tendon is nuch reduced. In the accipitrid the reduction is even greater, a connective tissue brace being absent but re- pLaced functionally by rnore diffuse connective tissue binding of the skin to the axilla. Iraneoeneue Perinei The traneversue perinei (fig. 186) is variabLe within groups

CAUDILIOFEMORALIS PUBIS TRANSVERSUS PERINEI LATERALISCAUDAE VENTRALIS PARSDEPRESSOR COCCYGIS PARSINFRACOCCYGIS

EPRESSORCAUDAE A B

Fig. 186. Ventral view of cloacal aperture area of A. Cathantes auna, B. Pandi,on haliaetue . particularly as to the site of origin and the size of the belly. The cathartid differs in that the main origin is fl.eshy from along the posterior margin of the tip of the pubis and the insertion is tendinous on the cloacal aperture nass. Sagittariue is like the accipitrid (including Pandion) and falconid in that the origin is usually from the posterior angle of the ilium above the eaudiliofemonaLie nuscle. The origin may be diffuse as in Gampsongcwhere it is more f,rom the general connective tissue in this region than fron any single point. Henpetotheres is sinilar. The falconids show about the same range of variation as the accipitrids. Depressor Caudae The depnessor eaudae shows some variation. In the cathartid (fig. L86) a separate slip inserts'on the cloacal sphincter mass: this slip is the division Leoatop eLoaeae. In Faleo, Buteo" and Sagittariue, a connection between the sphincter fascia and the nedial nargin of this muscle occurs but a separate "Levator eLoaeae" has not developed. Such a fascial connection allows both re- traction and levation of the aperture. Latenalis Caudae VentnaLi.s The pa?s depneeeor eoecygi,cis a separate nuscle in the falconid, accipitrid, and sagittariid; it is combined with the pars FALCONIFORMES 51(251) infra eoe ey s i s in the cathart il"f.)t";"lir);" In the cathartid, the pans stepnaLis of the stet,noconaeoid.eus arises fron the dorsal, anterior nargin of the sternum adjacent to the coracoid and extends as a thin sheet to the dorsal margin of the sternocoracoid fossa of the coracoid. In the falconid (and Sagittarius) the origin includes the anterior nargin of the sternocbracoid process of the sternum (fig. 189). The accipitrid and sagittariid differ in that the origin in the later includes the area of the sternocoracoid fossa of the sternum beneath the posterolateral part of the coracoid. The pars eoetalis occurs in alL falconiforms although- it is almost, if not entirely, lacking in Pandi.on and vestigial in Iorgoe. This part is relativel,y sma1l and arises fron the sur- fac6 of the sternal segnents.of the ribs just posterior to the sternocoracoid process-of the sternum. Its fibers pass forward and downward and insert by a distinct tendon on the truncated process coracoid in tltu s-agit-. tip of the sternocoracoid of the - tai'iid, accipitrid, and cathartid (fig. 1-91). Arnongthe falconids the iniertion is by a broad membranealong the rounded lateral margin of the coracoid. In Pandion or Torgos, thil insertion is largely or entirely taken over by the pars sternali,s. ?bliquus fnternus Abdominis The internal obliqu6 of Sagittar+us is peculiar in that the belly is restricted to the anteiior half of the abdonen whereas in the others it extends, oo its nedial margin, posteriorly to the pubis (fig. 187). Reetus Abdominis The reetue abdomi,nis is best developed in the cathartid (fig. 187). The belly extends posteriorly nearly to the- pubis ind-lateritfy it is continuous, or nearly continuous with the internal obliquus Transr)ersus Abd.ominie In the sagittariid and cathartid, the f ibers of this laye-r run nearly traisversely (fig. L87). This condition is approached in Pandion and the Pernin genera. Latiesimus DoYsi The Latiesimus dorsi is fairly uniforn throughout the falconiforns. Its pars anterior and pare posterior bands are most (cathartidt' and narrowest powerful in the iarge species -Torgos), imong the falconids- (fig. L90) . Asso_ciated with the posterior band-is the dermal pa"s Latiss'i,mus of the eutaneous matimus, lrlhen this cutaneous-nuscle is lacking, the posterior part of the- latissimus is closely bound to the skin. The over-al1 aspect of ittit region in Sagi,t'tanius is distinctive in that the very broad anterioi part of ihis muscle and the well-developed pars Latlssi- mus of th6 eutaneoue maa'Lmusnearly conceal the pars postenioz'. Teres Maior (proeeapuLohumeralis) Among the falconiforms, this muscle (fig- 188) is- perhaps (not absent as indicated by B918er best developed in Polthieraa ' 1956b) and Mieraetur. Polybonus has a thin round rnuscle like the aicipitrid or cathartid. Sagittanius is peculiar (Possibly an individual variation) in that this muscle is divided into two 52(252) M. .]OLLIE

Fig. L87. Ventrolateral view of abdorninal muscles, 1eg raised, of A. Cathantee auna, B. Sagittarius aerpentariue, C. Pandion haliaetus or Gampeonys suainsonni, D. Buteo iamicensis or Aeeipiter eooperii, E. Elanus eaeruleue, F. PoLyborus eheriuay or Faleo sparuerius. FALCONIFORMES 53(253)

PECTORALTS DELTOIDCLAVICULARIS CORACOBRACH]ALIS HUMEROCAPSULARHEAD LONGUS XERES MINORMRS EXT. rolD SCAPULARIS SUPRACORACOIDEUS

SCAPULARIS RES MAJOR

SUBCORACOSCAP. TrsstMUs mRsscAP.

EXT.HEAD OF T

Fig. 188. Dorsal view of head of the pa?s etternus of the triceps and adjacent structures in A. Sagi,ttari.us serpentarius, B. Coragypa atratue. separate parts (belLies and insertions), which lie close together. SubeoraeoaeapuLanis (includes dorsaLi,s seapulae) The pane eubeoyacoscapularie has three surface subdivisions: pa?B seapularie posterion, pars seapulari.s anterion, and pars eoraeoi.deue. The falcons differ in that the coracoid division is conpletely separated from the others at its origin and throughout _ most of its beffy (fig. 189); its origin extends down the shaft of the cor,acoid nedial to the insertion of the sternoeoraeoideus. Polyborue and Herpetotheres are like the cathartid in showing a massive, closely-connected, coracoid division, but in the cathartid the sternoeoraeoi,deue is concealed nedially (lies under the liganentous sheet attaching the coracoid to the sternal margin). The coracoid division in the accipitrid and Sagi.ttarius is nuch smaller and closer to the angle while the anterior and posterior scapular divisions are hardly separable. Sagittarlue has the snallest coracoid division and there is a large gap between 54(25\) M. JOI,LIE

Fig. 189. Medial view of the shoulder region of A. Buteo jamai.eeneis, B. Henpetothenes eaehinnans. its area of origin and that of the eoraeobraehialis breuie, The pays seapularis posterioy (donsalie-seapuLae) is constant in its form- and rel-ationships, although in Vultur the insertion tendon bifurcates slightly (Fisher, L946:586) . DeLtoi,deus The pa?s elaui.eularis gives rise to both the tenso_vP?tagi,i Longus anb tensor patagii bieui3 tendons (f ig. _1?O_l. In the falLoniforns, a diitinct slip fron the sternoeleidoecipi,taLi.s. is lacking aithough a vestigial one nay be found in the cervical patagium. A slip-fron the peetorali,s is commonto all. The rbeneon i,att6r.ns of the patagii tendons is somewhat distinctive. The brevis tendon of Sagittaniue has an outer accessorl division which separates before ttre midpoint and joins a brace frorn the longus t-endon. Al,so the brevis tendon spreads out distally-so ttt.i the insertion is rnore diffuse and with less well-marked braces. A connective between longus and brevis tendons also occurs in the cathartids, Gypa'dtue' Pandion, and Gypohie?at; it is indicated in Torgoe an

Fig. 190. Dorsal view of shoulder rnuscles and tensor patagii area of-wing in A. Sagittanius serpentarius, B. Cathartes aunar, C. Pandi.oi haLiaetu-s, D. Buteo iamai.censis, E. Faleo spanuerius, F. Polyborus ehertuay. 56(256) M. JOLLIE tinct in its niddle section. The pars seapulanie of the deltoideus distinguishes the cathartid- in that the origins of the two heads are continuous and the oe humeroeapeulanie is lacking. The sagittariid, _accipitrid, and falconid have two distinct heads, the deeper one from an oa humeroeapaularia (fig. 188), ruo"" Minor The pane ecternue (fig. 188) is well developed in. the cathartid- and arises along the outer aspect of the tip of the head of the clavicle; its deeper division arises fron the acro- nion process and the anterior, outer nargin of the scapula_. These two divisions insert separately in Cathartee (slightly di- fused), Gymnoglpe,-or and Vultur. In Saneonamphue,th9 deeper visioir is- more- Less lacking . (Fregata lacks it. ) Only thg deeper division appears to be present in the sagittaliid, qgql- pitiid, and falcoir-ia. The accipitrid, including Pandionr- differ in ttrai. the be1ly inserts along the posterior nargin of the tendon of the supraeo?aeoi,deus-for most of its length. In the sagittariid and ialconid, the belly is free and its insertion is coifined to the point of- insertion- of the supracoracoideus and beyond.' The pars internus occurs only in the cathartid. It arises from the iranubrial region of the sternun and the nedial- edge of the coracoid; the belly overlies the rnedial aspect of the sup?aeo?aeoideus and extends out through the triosseal canal afbng with that muscle. The broad tendon of insertion overlies that-of the supraeoraeoideus and attaches in comnonwith the p,ars enternus. Thii part is innervated along with the externus and is thus not a part ilf the eup"acoraeoi.deus (see Fi.irbringer, L902, footnote p. 535). Inieepe Braehii pa?g The triceps of the cathartid differs in that the -e8- ternus (long hbad) has two separate and distinct tendons of origin, one at right angles to the other (fig.188). The first coniinues the axial line of the belly and attaches to the scapula glenoid- prbcess; this is main and iu-t behind and above the -the iypical origin. The second is a brace which extends nedially f?-omthe anterior end of the be1ly to attach to the dorsal aspect to the scapula below the anterior margin of the pars seapulanis poeterior bf the subeonaeoseapularie. In this group the brace to the humerus is broad and strong. The scapular origin is widest' and entirely tendinous in Sagittari.us, but ther6 is a broad brace suggesting- the cathartid. In"the other falconiforns, the glenoid origin is the-only one and consists of a strong round tendon,(as in the cathartid) and-dor- sonediaLLy there rnay be a thin tendinal or fleshy goltinuation.* * * rt * *-* * * * *-* * * * * * * * * rt * * * * * * * * * * * * * * Beddard (1898;474) comrnentedon the variations of origin of the pane eaternus, "The anconaeue arises in.Polyboroi'.des l#ymnogenysT and in some other typgs by a single head from the scipuli, ihich is partly fleshy and partly tendinous. In Vultur tr fbrgoiT aurieulaita, bn the other hand, the muscle arises by two c6mpletely tendinous heads, so that the nuscle has not that FALCONIFORMES 57(257 )

The muscle of the accipitrid and falconid are nuch the same except that in the forner the origin tends to be broadly tendinous and a humeral brace is better developed. Further distaLly a broad connective tissue band braces the belly of this part to the hunerus in alL. The pane internue is well developed in a1.L1:€uld a pa?g asillarie is present but sometimes extremely rninute in the falconid and cathartid (fig. 190). This part arises off the tendinous band of the enpanso" seeundari,o?um. Peetopalia The peetoral;is of the cathartids'is completely divided into a pa?e eupenfi,eiali,e and a pa"e profundue. In the other groups the belly nay show some division but the layers do not seem to correspond to the parts of the cathartid. Sagittani.ue (fig. 191)

BICEPS

PECTORALIS

CORACOBRACH. ERRATUS LONGUS

AXILLARY CONN.TISSUE

TERNOCORACOID. PARSCOSTATIS CORACOBRAC EREVIS SUPRACORACOIDEU

Fig. 19L. Lateral view of body, wing raised, and pectoralis muicle removed of Sagittariue se"pentarius. Origin and insertion of the pectoralis are cross hatched. titt*** * **** ** *** ** * ** **** ** * *** * value in the classification of the Aeeipitres that I at one tine thought. " 5B(258) M. JOLTIE

has a greater area of origin for this muscle beyond the margin of the sternum than any other falconiform, this correlates with the shortening of this bone. A pa?B patagialie occurs in all but presents no useable informat |o*2;"r""ot)d,eus The I upr aeo r aeo i. d,e ue i:" i;;"g;!i" i i" .r,u ca thart id (exc lud i ng leratorni,e--fig. 191), its origin extending nearly to the posterior end of the sternun (fig. 160). In the accipitrid, falconid, Sagittarius, and Teratornis, it is Linited to .the anterior half. Corae obraehi.ali, e Br evi e In the falconid, the eoraeobraehialis brepi,e (postenion) arises off the coracoid and an adjacent area of the sternum. As a result the tip of the sternocoricoid process of the coracoid is concealed whereas in the other faLconiforms it is exposed and braced by a strong ligament to the sternun (fig. 191). Bieeps Braehi.i The bi.eeps of Sagi,ttarius is different in that the bellies of the coracoid Lnd huneial origins are slightly separated and soon fuse (fig. 191). The insertion tendon gradually divides and attaches to both the radius and ulna. The insertion tendon shows a slight subdivision in some accipitrids (aiLuus, Pandion). Ieneon Faee'ia Latae In the cathartid, the teneor faxeia Latae arises along the entire il"iac crest fron the posterior nargin of the sartorius to near the posterior end of the iliun. The belly, near its in- 'sertion, is braced to the outer aspect of the shank by a fascial binding. The preacetabular part inserts by an aponeuroses over the uastus Lateralie as in the oLher falconiforms while the postacetabular part, thick and fleshy in Coragype and Saveoramphusand proportionaLly snaller in Cathartes and GymnogAps,inserts along the posterior margin of the Dastus Lateralie and the anterior part. The other falconiforns lack the postacetabular division of this muscle. In Sagdttarius the two slips indicated in the accipitrid or falconid muscle are separated (fig. L92). Herpe- totheres and Mi,eraetun approach Sagi.ttariue in differentiation of these s1ips. ILiofibuLanie The anterior proximal nargin of this muscLe is not fused to the tensor faseia sheet in the cathartid as it is in the others (fig ' 192) ' Gluteus Marimue (pini,fonnie) Pandion is peculiar anong the falconiforms in that the gLuteus macirnus is fused with the gluteue medius (gluteus profundus). Hudson (1948:L03) did not record this fusion for his specimen. It may be that there is some individual variation in the species. Pini,formie (iLiotroehanterieus nediue) The falconiforms can be divided into two groups on the basis of the fusion of the gluteue ninimus and pi.riformis (fig. L92). In the accipitrid, Pandion" falconid, and Sagittarius these muscles are nearly or conpletely fused, and the line of fusion is indicated by the point of penetration of the superior gluteal nerve. In the cathartid these two muscles are separated and distinct 4S, for example, in the crow. FATCONIFORMES 59(259)

IBULARIS GLUTEUSMEOIUS ICEPSFEMORIS ET SEMIMEMBRANOSUS roR coccYcls ERALISCAUOAE DORSALIS LATERALISCAUDAE VENTRALIS

SARTORTIUS

MAXIMUS

TENSORF LATAE

RECTUSFEMORIS TENDON

EMORALIS VASTUSLATERALIS FLEXORDIGITORUM SUPERE STROCNEMIUS GASTRbCNEMIUS

VASTUSINTERME FLEX. HALL. LONGUS PERE DIG. IV I ET PERE DIGII PERONAEUSLONGUS LEX. DIG. PROF.

TIBIALISANTER TIBIA

BET PERE DIG. III PERE OIG. I B PERF:DIG. II

Fig. L92. Lateral view of thigh and upper shank of Sagittarius serpentatiue. A. External layer, B. deeper layer (extenson di.gitorum Longus not shown). 60(260) M. JOLLTE

Reetue Fenori,s (ambiene) The belly of the neetue fenonie (fie. 193) is present and

VASTUSLATERALIS

R€CTUSFEMORIS

SEMITENOINOSUS

VASTUSMEOIUS

PLANTARISPARS ACCESSORIU

PARSTIBIALI

GASTROCNEMIUS

PLANTARIS PERONAEUSL SARTORIUS

Fig. 193. Inner aspect of thigh of A. Sagittarius serpentarlus, B. Pandion haLiaetue,

constant in its form in all of the falconiforms with the exception of the pernin kites (only Aoiceda and Chondrohierat dissected). Here, it is lacki-ng although the tendon is present in the shank. In Chondrohienar the tendon arises deep to, and from, the fibular attachment of the iliofibularis frenula or the lateral condyle of the femur and acts as an origin for the f'Lenor digitorum supe?- fieiaZi'a muscle. In Aoiceda, it acts nainly as an origin for the pars perfonatus digiti ff as in the typical accipitrid. In these kites, the remnant of the tendon passes lateral to the insertion of the i.Liofibularie. Elanus (and Gampsonyx) is the only accipitrid in which this nuscle is fully developed., Yet the tendon passes lateral to the iliofibulari.a insertion and acts as a general origin for the entire conplex of the superficial digital flexor. Pandion has a relatively broader reetue than rnost accipitrids. The tendon passes rnedial to the iliofibuLaz,is insertion FALCONIFORMES 61(261) and acts as the only medial origin for the superficial flexor conplex The falconid differs fron the other falconiforms (with the exceptions cited above) in that the tendon of this muscle usually passes lateral to the iLi.ofi.buLarie inseltion- (Hudson, 1937:16; fS+A:104) and serves as a commonorigin for the parts of the superficial digital flexor (as in ELanue). (The tendon ulYl}ly enis on the belly of the pane perfoncttua digiti. rII.) Polihieran is an exception (Berger, 1956b:520); here the tendon passes medial to the ili.bfibularie-insertion. Sagitta?ius and the cathartid are like the typicaL accipitrid. Vastue Lateralia In the falconiforms, the uaetus i,n'benmediuaand Dastus Later- alie are not separable. The uaetus Lateralie is sonewhat divisible into outer and inner parts at. the point of penetration of the nerve (fig. 193A). Here, superficial Layers of muscle fibers are evident (the division lateralis), but, when seen in section, lhere is no separating membranepresent. The insertion tendon of the internal- division is poorly rnarked in accipitrids, falconids, and cathartids; it is stronger in Pandion and well rnarked in Sagi.ttariue, Vastue Mediue (fenoni.ti.biali'e intennus) There is some variation in the form of this muscle in the accipitrids. In Buteo it appears to have two parts _whereas in most genera only one part is evident. Neeroeyntes is sinilar to Buteof here, the proxirnal end of the muscle is parti?LLY divided. Aquila showi onLy- a surface tendon in the position.of the tendons oi Buteo whereas-Pandion, Ciyeus, Aeei,piten" f eti'nia, ELanue' . Aoieeila, and the falconid genera have b thin surface tascia, which condensbs distally and along the posterior margin to forn a part insertion tendon., This typg- of muscle superficial -only of the corresponds to the rnedial division of. Buteo. The muscle of pandion has a median superficial fissure that does not correlate with any deep tendon and its significance is-questionable. In the belly and convgTgg to Sagittariua several bundles form -the iniertion (fig. L95A). Cathartee shows distinct medial and internal divisidnsl Uut Conagyps has only the central tendon of two such (unseparated) parts. The-internal division of Cathartes is very imal1 conparea-wittr the muscle of Buteo but otherwise resembles it. The variations of the uastus mediue appear to have little, if any, taxonomic vaLue and cannot be correlated with size or function. Vastus fntermedi'ue sagittariue is unique in that there is a separate and distinct uaetue intenmedius (fig. 1928). Bieeps Femorie et Semimembnanoeue The pars Lateralib (bieepe femoris) is present in Sagittari.us (fig. L92) and the cathartid,-it is-nissing in the-accipitrid ald iallonid.' This muscle is associated with a part of the pLanbaris (the accessory head) and is further identified by_-its origin and its position lateral to the nerve of the eemitendinoeue. The insertion of this muscle on the "accessory" is characteristic (except in Dnyobates pubeeeens, a woodpecker, where it inserts like 62(262) M. JOLLIE the -Thepars medialis). pa?s nedialis (eemi.membranosus)is found only in-the falconid.- It arises in comnonwith the semi.tendinosus and lies nedial to the nerve of that nuscle, (Faleo, fenaeidea, artd poLi- hieran) or the nerve passes through its belly. It is further identified by its insertion near the eemitendinoeue. SenitendinosuB Sagittari,us differs in that the semitendinoeue inserts nearer the head of the tibia than in the other types.' Gtaeilie et Adduetot Longua The cathartid differs from the others in that this superficial adductor layer, the gnaei.Lis et addueton Longusis fusedr more or Less in different genera, to the deeper adductor except near the origin of the pane addueton Longus. The pane graeilis is indistiiguishable (weak line of separation in Coraggps and Cymnogype)fron the posterior nargin of the deeper layer itt TI " slecirnbns, but Fisher (1946:668) described i! (also Hqdson' 1948: 1b6) as distinct. Fisher's description of the form of the pa"s gnaeili.e as "A thick spindle-shaped posteriof Part'f _is not ap- ificable to ny specinens. In most acgipitrids and falconids the paoe gnaeili,s' is'distinguishable by tfte length of i!t fibers and on the origin of the pTantarie ieparite insertion -(fi9..1931): In' Sagittaniue, the cathartida, some accl.pitrids (fetinia" Aui'eeda, glanu:s, Neenoeyrtes) and some falconids (poLybonus) it is, at best, onLy poorly marked Adduetor Brevi.a et Magnue The ad,d,uetorbretsi.e et magnus of the cathartids tends to be fused with the superficiaL adductor layer. In the others this layer is free. The muscl"e of Pandion arises-by a.relatively narrow, strap-Like tendon just anterior to the origin of the semi*natnoeis. The belly- fans out fron this origin to its thick and fleshy insertion along the distal third of the shaft of the fenur. I; the renainder of the falconiforms, the muscle arises along the ventral edge of the ischiurn for some clistance and inserts as in PandLon. Abbunator Eaternus The obtunator enternus is constant throughout the falconiforms with the exception of Sagittarius where the parg internus arises from the innei surface ol the posterior, lateral part of the iliun, as well as from the inner aspect of the ischium and pubis. A situation similar to Sagittaii,ue is found in the chicken. In the this muscle large cathartids (Sareonamphue,'dorslt GymnogU?s,Vultut), has two tendons, ih" more one 6eing the longest; these fuse at the distal end of the belly to insert as one. The dorsal 1"11y of the pans erternua is lackiirg_gnl-y in Pandion (Hudson, 1948:106). ttf-,2d"":"":rt\i""TZk3"r"ru" rhe pa?s eaud.atua$ i-s absentin Sagittariue, GymnogUPs,Sdreoramph-ue, ald Vultur but present in Coiagype and Cathar+ee and other- falconiforms. Both the origin and iiriertion are tendinous when this muscle is present in the cathartid. Its loss in the condors is an indication of their specialization from the ancestral condition shown by the smaller species. Among the accipitrids the origin is usYalty by ?_short fiat tendon and the insertion is tapered and tendinous. However, FALCONIFORMES 53(263) the length and width of the origin tendon varies as does that of the insertion. In Pandi.on, Iorgos and Neeroeyntee, the insertion is by a Long and thin tendon. In Cireus and retinia it is broad and fleshy. Miloue shows a nore or less internediate condition. The falconid resenbles the small cathartids and some of the accipitrids in having a long thin tendon of origin and a short, fairLy broad insertion tendon . Sagi,ttariue is the onl.y falconiforn which has the pane i.Liacue (fig. 1928). Caatnocnemius The pattern of the para Lateralis and pard medialie of the gastroenemius (fig. 194), as well as that of the plantaris and eoleus, differentiates the sagittariid and. cathartid from ,the other falconiforms. The cathartid and Sagi.ttanius agree in having the medial part arise in part from the patellar fascia. Sagittariue resenbles the accipitrid in having strong tendons for both parts, These tendons are just distinguishable fron the generll sheet in the cathartid. In the falconid the nedial tendon is only slightly marked (strongest in Henpetothenee and Mienastun), PLantarie The cathartid and Sagi,ttari,ue are distinctive in having both a pars aeeeesonius and a pars tibialie. In the cathartid, .the pals tibi.alie inserts with the lateral part of the gaetroenemius rather than separateLy or with the nedial part. The falconid and accipitrid are similar. SoLeus The soleus is lacking Ln Sagi.ttarius (and Chondrohi.era&among the accipitrids). It is a separate entity only in thg falconid and cathirtid. In the cathartid, the belly is very short (less than one-fourth the length of the tibia), and the tendon is long and slirn. In the falconid, the be1ly may extend two-thirds or more of the length of the tibia and is partly exposed posteriorly below the margin-of the medial part of the gastroenemius. The extreme in the falconid group is found in Henpetotheres and ltieraetur where the distal part of the belly is fused with the nedial part of the gaetroeneniue; the insertion tendon is separate. In PoLi.hie-rax, the belly has a snall slip which inserts with the plantanie. The muscle otherwise is like that of Faleo (may be in- iividuaL variation--not reported by Berger, L956b). In the accipitrid, the eoleue is always vestigial and fused distally with the gastroenemius. It is lost in Chondrohi,elaa. The proxirnal end 6f tne belly (lateral to the insertion of the eemitendiri,osua) is most distinct in some of the kites and vultures and in Pandion (fig. 1938). It was also well developed in a specirnen of Aeeipiter striatue. rhefL e ror oonul3""i,x" f,:!u"t\x|uf in birds (fig. 1928) which share common""3z",ix2i2ti!'origins but which usually"'., ins-ert sepirately, across the heel and down the posterior aspect of the_ taisonetatarsus r otr the toes. The configuration of the parts of this muscle distinguish each of the groups of the falconiforms. The pane perforane et perforatue di.giti, rr,lies gntirely ex-- posed lat-eralIy in the falconid; its origin is from the rnargin of ttre ti.bialie antenion (or pe?onaeus Longus) proximally and fron 64(26\) M. JOLLIE

MITENDINOSUS

J = E, m lrl 9 F D J r

A B c D E

iiitrr' ' i::i:,j:!!

F G H J

Fig. 1g4. Senidiagranmatic dorsoposterior views of gastrocnenius- plintaris pattern in A. Cathartei auna, B. Sagittanius eerpentayius, b. Pandi.on- haliaetus, D. Elanus eaeruleue, E. Ietinia plumbea, .F ' Mi.Lous migrans, G. Neeroeyrtes monachue, H. Mi'cnaatur nufieoLlie ot He"pet-otherLa eachi.nnane, I . Polihieran eemitorquatue, J' Polgborue eheriuay. FALCONIFORMES 65(265) the fibula dista1ly. The other groups agree in having the anterior origin fron the nargin of the perfo"ans et perforatus digiti fff (the reetae femoris tendon possibly involved here in most falconids and some accipitrids) and the belLy is partly or entireLy (in Sagittariue) concealed by the lateral part of the gaetroenemiue. In Pandion the tendon of this part and that of the pane penforatus dngifi. II fuse in the calcaneaL region, ald throughout their tarsal length, and insert in c,omnonon the base of digit II. Correlated with its restricted function, the belly of this muscle Ln Pandion is proportionall.y snallest. The pare petfonane et penforatue di,giti IIf of the falconid is concealed by the previous part, In most of this group (except Polihienao) and in some accipitrids (Auideda" Chondrohierax, Elanue, Gampeonyo)it arises in part from the reetue femonis tendon along wittr the pana perfonane. et perfonatue 4yglti fI and the pans perf6ratue di.gnVi IV). The cathartid has a distinct fibular origin -afoirg the line of contact with the ti,bialie anteri,or whereas the othei groups have little if any direct origin-from that bone, except at ttre extreme proxirnal end. The origin, frorn along the. margin of the ti.bi,ali,a antenior (or peronaeus Lo-ngue), is restricted 5r even lacking in the accipitrid- but extensive in the falconid and in Sagittariue, The lane perfordtue di.gi.ti .flr has both lateral and rnedial origins, bne oi the other of which nay be dev-elope.d.to the near exclusi6n of the other. The medial head of the cathartid is large and fleshy (arises by a very short round tendon) whereas the lateral hea'd is a broad aponeurosis as weLL as an attachnent to the bellies of the previous parts (in Cathartes and-Coragyps)-. fleshy fron both the Lateral Sagittariue differs in- having -heads head and a tendinous, or largely tendinous, medial one. Gampsonyn and ELanua lack a distinct lateral head and have a large fleshy rnedial one for this part. Pandion is peculiar in lacking -a distinct nedial tendon; this origin is entirely replaced by th9 neetus fenotie tendon. The filconid differs fron the accipitrid in that the lateral head arises near or at the proxinal end of the shank and, in part, frorn the reetus fenonis tendon (see abooe) whereas in th; latier itre aponeurosis of origin extends to the niddle of the belly of the pans perforane et perfonatus di'giti' II ' In those accipitrids in whith the reetus tendon passes lateral to the il|ofibularie insertion (eee aboue), this tendon serves as the only latlral origin. The medial head is tendinous in the falconid, or, as in Polgboiue, is represelited by a few, fleshy fibers from the nedial teirdon which also acts as a part of the origin of the -pa"s perfonatus di,giti rIr. ?he pa?s perfonatua digiti III typically has two separate heads of brigih. In Elanzs and Gampaonyx,thes-e are conpletely separated anf, their tendons do not tuse until_ they pass over the caicaneal mass. A strong vinculun connects the tendon of this part with that of the pars perfonanl e! perforatue- di.gifi Jf-f near tfte distal end of the tarsometatarsus in Sagittarius and the cathartid. In VuLtur (Fisher, L9462679), a vinculum occurs between the tendon of this belly and that of the pa?s perforatus di'giti. IV. The pa?s penforatue digi.fi .I.r in the cathartid and Sagittarius 66(266) M. JOLLIE

arises nainly from the medial origin fron below the belly of thg irl the accipitrid, Pandion' a4d par perforatus digi.ti IV -whereas fafc'onid the lateial head gives direct origin, The reetus tendon acts as a secondary origin in each instance; it is more involved in the accipitrid and cathartid and Less involved in the falconid and Sagittanius, Tire relative positions of the tendons in the calcaneal re-gion is constant in all-nernbers exanined, the flexors for digit II being deep and somewhat separate from the other tendons. Th9 genus palbo has an ossified tibial cartilage lying over the hypotarsus in this area (see fig. t43), which is lacking in the other falconids. The insertions of the various parts of this complex are variable (fig. 195) but do not show any correlation as to grgup. Each-of-the several groups of the falconiforms can be characterized on the basis of tf,is muscle and each type appears to be completely independent of the others. The greatest variation is can as touird anoirg accipitrids with which Pandion !9 -associated the most a6errant style. ELanus and Gampsonyxdiffer in nany ways from the other accipitrids and show some sirnilarity to the pernins (Chondrohierar and Auieeda) . Fleson HaLLueis Longus Longus differs in Pandion The origin of the flecon haLlueis - from that of-the accipitrid and falconid in that it arises laterally by a powerful tendbn. The cathartid differs in that there is b6th'a llteral and medial origin from the fenoral condyles between which the femoral origin of tf,e superficial flexors arises (pats- piifo""t"" di.gi.ti ID: The double area of origin is-higher on.the ?"*Lt in Sagiitanir,rs than in the other groups.. In llt" cathartid, the two areas of origin are continuous below the rnedial head of the smaller iiLto" digntorum comllex. The belLy of this muscle is nuch it the caihartid thair in the other falconiforms (less than one-half ifiu length of the tibia), and it is largely concealed by the superficial flexor mass. ' The nature of the vinculum between the tendons of the f,Lqqqf (1898;^ hal7ueis and,fLeror d.igitonum Longus is described by-Beddard i?;i""i-oi--ittu" types, one of which, ascribed to Auieeda (Baza).,-. second digit. This genus-(fig. lacrts a recognizable division to the - 1g6C) is pec[liar in that instead of a single branch-, serving digit iI;-pa-si^ng through the vinculum, lwo distinct branches occur. The f iist of tfiese seives digi ts I I and I I I and the second' lyir-tg above the first, serves digits III and IV. In no other accipitrid examined is there a diitinct branch to digits III and IV. Chond.nohienaais peculiar in that it shows no branches to the front digits associatei with the vinculum, or els^ewhere, and thus ,n"t.h"t ileddard t s third type. The concurrence of vinculum and branch to digit II is not-limited to Auieeda but occurs in most accipitrids. Torgos is peculiar in the length of !h" vinculum whicfi extend.s neaily one- fifth the tarsal length of the deep flexor tendons. In general the insertion of this muscle in the falconid is like that of the accipitrid except that th9 tendons are ossified. Pandion and the cathaitid are similar in that the tendon of the fLeror haLlueis fuses for nearly half its tarsal length with that FALCONIFORMES 67(267 )

ffi ffiffiffiffiffiffi

ill

ffiiinNl ffi I u'l mffiffifi fi fi IV dh ffi ffiNffi|1Ji A A B ffi C ffi M F D ffi E G H ffiI ffi K

Fig. 195. Ventral view (semidiagrammatic) of insertions of tendons of the superficial flexor compl-ex on the phalanges of the toes of A. Sagittarius eenpentariue, B. Cathantes aura, C. Pandion hali.aetue, D. Elanus eaenuleue, E. Atticeda eubcrietata, F. Chon- dnohierac uneinatue, G. Cineus eyaneua, H. Buteo ianaieensi's, I. Necnosyntee monaehus, J. Faleo sparoeriue, K. PoLybonus eheniuay 58(268) M. JOLLIE

FLEX,HALL. AA A AA A B C D

E G

Fig. 196. Deep flexor tendons, viewed from in front (left) and Ue[ina (right) of A. Coragyps atratus, B. Sagittarius serpentanius" C. Aoieeda subcnistata, D. Chondrohieran uneinatus, E, Buteo jamaicensis, F. Pandion haliaetus, G. FaLeo sparuenius, of the fLecor digitorum Longus so that fibers from this muscle serve aif four t6es (as they do to a lesser extent in Auieeda). As a result of the relationship between these two deep tendons, the fleaon hallueis of. Pandion or the cathartid lies somewhat more lateially in the calcaneal region than in other falconiforms. In Sagittanius, these tendons blend just above the digital branches. Flenor Digitonum Longus The relative size of the belly of this muscle varies in different groups, t.hat of the cathartid is smallest. In the accipitrid ind Pandion, it is usually exposed on the lateral-aspect bf tne shank between the peronaeus Lbngus and the superficial parts of the fleror digitbrum superficialis. In Auieeda, it is irearly concealed whereas in ELanus it is very prominent. This muscle is exposed only at its extreme posterior end in the other FALCONIFORMES 69(269) groups due to variations in size and interrelationships of the clif ferent muscles. Tibi.ali.e Anterior In the accipitrid, the tibiaLis anterior is a superficial muscLe of the shank whereas in the cathartid and Sagittarius it is covered by the pe?onaeus Longus. The falconld may have this muscle covered (Polyborinae, Micnaster) or entirely exposed. When exposed the lateral rnargin is bound to the nargin of the fleaor digi.torum superfieialie and the underlying pe"onaeus Longus. In tha cathartid, where this rnuscle is covered, the lateral line of attachnent is reduced (eolybonue is like the cathartid). Like the others, the cathartid has a medial membranousorigin fron along the tibia (contrary to Hudson, L948:L06), but it is reduced and lresent only distally. Sagittari,us has a similar origin medial to the ertensor digitorum Longus. The insertion tendoh of the cathartid and sagittariid bifurcates to attach to the two closely associated tubercles on the anterior aspect of the proximal end of the tarsometatarSus. That of the othei falconiforlns is undivided and attaches to the single tubercle. The accipitrid (not Pandion) and some falconids (not Polyborinae) have a secondary tendon (or two) inserting into the surface fascia of the tarsornetatarsus. Peronaeus Longue The peronaeus Longus covers the lateroanterior aspect- of the shank in ihe cathartid, sagittariid, and some falconids (aLso Neophnonaccording to Berlin, l-963, J. Animal Morph.- and PhI?. g:fi). It is nuch reduced in size in the others and is lacking only in Pandi,on. In Faleo, Ieraeidea, PoLihie?an, .and.He"petotheres, this muscle is like that of the accipitrid; in Mienastun, it has a strong aponeurotic origin which covers the tibiatis anterion and in MiLoago and the other Polyborinae muscle fibers cover much of that muscle. The forrn of this muscle in the cathartid, sagittariid, and the Polyborinae appears to be the prinitive type (occurs in most birds--iludson, L93i:35). Reduction of this muscle is correlated with a nore powerful cienching foot. The extreme is Pandi.on and the owl in which it is lacking. Peronaeus Breuis In pandion, the origin of the peronaeua brepis extends proximally, as conpared with the others, to include the greater part of the length of the fibula. Extensor Digi,toYum Longue Al.though the nuscle itself ihows no real variation the frenula associated with its tendon does. It is on the joint or the rirn of the tarsometatarsus in the falconid, whereas in the others it is on the proxinal end of the tarsometatarsus. In Pandion, this frenula is ossified. In most species, it is a distinct fibrous band, but in Iongoe and Sagittanius it is a 1ong, indistinct, enclosing sheath. In Cathantee, it is a very narrow band set in a well-developed sheath. The distribution of the extensor tendons to the toes is quite irregular. In Cathartes and Coragyps, the tendon bifurcates- just above the third trochlea and the branches extend down onto the 7o(27 o) M. JOLLIE second and third trochleae. The lateral branch is the wider, and, between the third and fourth trochletat,it divides into ten- doni which go to these digits. The main tendon of the accipitrid b ifurcates one fifth of the length of the tarsometatarsus above the toes; the tendon for digit III tends to arise from fibers from the lateral division (fron the medial division in Cireus) but becones distinct only at the base of the third digit within the fibrous connecting sheet between the two prinary divisions (also in Sagittarius). Pandion ald Torgos have distinct branches irom both the medial-and Lateral divisions contributing to the tendon of the third toe. The falconid is like the accipitrid; onLy Polyborus differs in that the tendon for digi! III arises largely irorn the medial division (but receives a distinct contribution fron the lateral one). Eatensor Digitorum Breui,e The pars hallueie of the et-tensoT 6n9?torum breoi,s (figs. hallueie) is sirnpl":! in the cathar- 1"26, L28'sagitianius,r' 130, 141" eatensor iia, or where it arises along the- line of fusion of netitarsali II and ift for most of their length' In both, the origin extends across the face of netatarsal II to ? posigiol aboie the first netatarsal. In Sagittanius, there is a slight- bifurcation of the proximal end of-the muscle to either side of the ti,bi,alis anterior insertion. This muscle is most variable in the accipitrid where it arises fron several heads. The medial head, usually the larges-t , includes the proximal rirn of the nedioanterior face of the shaft of the tarsonbtatarsus, the frenula of the estensoz' digitorum second Longus and a variable _length of the shaft below these. A heab, lateral to the firsi, attaches to the anterior ligament of the libiotarsal-tarsometatarsal joint and the general fascia of this joint. This niddle head may fuse-, in the area below the erteniop d,igitorum Longus frenula, with the medial head; it-is missing as i aistinct entity in Torgos r-althougtl- a bundle of fibers-arises fron the shaft in its- position (like the falconid) . The middle and medial heads are grooved for the passage of thq (slight in Pandion and nil.n"oo digitorum Longus tendon -groove the medial iread may overlap that tendon (AquiLa, Buteo, Aeeipi,ln"- Circua). LateraL io the tTUiatts anterior insertion, a third head (lacking in Tongos) from the- same general region as the ^"ysetond. "ti.r" Tliis head nay be- large or small depending on the presence of a fourth head fron the lateral rin of the tarsonetatarsus above fV. last the origin of the pa"s adduetor'estensor -digit-i _This head is"lacking in'Aceipiter striatus and Pandion. The various bellies unite distat to the tibiaLis anterior tubercle and the thick wide tendon extends down around the nedial margin of the tarsometatarsus to the upper surface of the hallux where it inserts upon the ungual Phalanx. Tha falconid-rnuscie has two heads (corresponding to the first and fourth of the accipitrid), both of which arise below the pioxirnal rim, on eithei side of the ti,bi,alis anterior tendon, from a variable length of the shaft (matched by Torgos). - The lateral head is usuall| the larger. These heads unite at the level of the tibi.aLis antenior tubercle and taper down to a tendon &t, or FALCONIFORMES TT(27T) above. the leveL of the niddle of the tarsometatarsus. it night be assumed that the nuscle of the accipitrid is the more specialized in that its origin has crept upward on to the joint and several distinct heads have appeared. The pars abductot digi.ti. rI may arise either proxinally' distally, or fron both areas within a group. Anong the cathartids, its origin is distal. It is largest irt Sareoramphue, arising as far 1atera1Ly as the face of metatarsal III. In vultur, the muscle has two heads; one fron the face of the tarsus and a smaller second head from the first metatarsal. Such a state is indicated in Cathartee, In Sagittavius, the origin is distal, from the anterior face, and bindings, of the first netatarsal and the adjacent part of netatarsaL ll. Among the accipitrids, this nuscle usually arises proxinally to either side of the tibialis anterior insertion. The lateral origin is restricted to the proxinaL margin whereas the nedial one extends down the face of the tarsometatarsus past the fib- ialis antepion tubercle--including the nediaL and distal aspects of this tubercle--for a varying distance (just past it in Cireue or Aeeipiter and nearly to the proxirnal end of the first metatarsaL in Aquila and Buteo). Aqui.La has both origins well developed; Buteo lacks the lateral one and the nedial one extends nore pioxinally. Cireue has both heads well developed although the l-ateral one connects membranousl-yto the medial one beyond the tibialie anterion tUbercle. Aeeipiter has the lateral head reduced to a narrow, fleshy band and the nedial one restricted to an area extending from the level of the tibiali,s anterior tubercle to a short distance distal to this tubercle. There is no distinct proxinal origin in fetu)nia or Aui'eeda. The nuscle arises along a line frorn the tibi.alie anterior nearly' to the distal end of the bone and from the adjacent nargin of the first netatarsal. In Torgos and Neenosyntea, only the 4it- tal origin is present, including a part of the surface of the first metatarsal. In Pandion, the muscle is even more distally res tr icted. In Mierastur, this nuscle extends much of the length of the tarsometatarsus and has both proxinal and distal origins which are slightly separated. The first extends for one half the length of the iarsbnetatarsus, its proximal portion underlying the medial head of the ertensor halluci.e. The distal beLly is very small and does not extend onto the first metatarsal. In sharp contrast' Herpetotheree has a powerful distal part which does include the antlrior nargin of the first netatarsal in its origin. In the other falconids, only the distal part occurs; it is very srnall, but incLudes the margin of the firqt metatarsal in its origin. The pans estensor di.giti fff is present in all. It is very small in bagittaniue and the falconid, arising just above the middle trochlea. In the cathartid, it is also minute but Gymnoggpsand Sareonamphue(Fisher, 1946:688) have an additional slnal1-tendon from a proximal origin. In the accipitrid, this muscle ranges from sonewhat better developed than in the cathartid (Aceipiter, Cineua" Buteo, Tongos) to strongly developed, the origin extending up the shaft to the tibialis anteni.or insertion 72(272) M. JOLLIE

or beyond (aquila, Pandi.on, Aoi.eeda, Chondrohieraa, ELanus, Ganpeonyn). The belly may show some subdivision. There are two heads in Pandion but these do not correspond to those of Aquila where both arise along a line on the shaft up to the tibialis anterior tubercle rather than side by side. Chondrohienan shows a partial longitudinal division, deep fibers arising separately wh-ich soon join the more superficial and medial belly. The pars adductor-estensor digi,ti fV of the Cathartid arises along the-line of fusion of metatarsals III and IV, fron the level of the tibialis antieus tubercle to just above the distal foramen. The origin expands abrupt dista1ly, fibers arising even from the face of metatarsal II. There is little variation among the cathartids, but the muscle is least developed in the condors. The muscle of Sagittarius is weakly developed. It arises by two thin heads which are confined'to the distal third of the long tarsometatarsus. The accipitrid and fal"conid are sinilar in that this muscle extends the length of the tarsometatarsus. It arises lateral to or underneath the belly of the ertenso" haltuei.e Longue for a varying distance. The insertion tendon, like the belly, rnay be very long or short Interossei Plantaris The pars abdueton digiti fV of the cathartid is well- developed'both proximaLly- and distally but is not two _parted. The point bf insertion varies from the margin to a short distance out bn the first phalanx (basal phalanx). In Sagittarius, there are proximal and distal bellies joined by ? 19ng t-endon_(Hudson, ig+g:108, described only the distal be1ly). The falconid is like the accipitrid but always has a partial origin-fron the calcaneal fasLia and usually a shorter belly (restricted to or Iorgosr exteTds proxinal third).' In Herpetotheree -the-origin irearly the length of the-bone. Where the tendon is long (acci- pitriS, falconia), it is closely bound to the lateral n?rgin of itre tarsometatarsus above the trochlea. The insertion is on proxinal of the basal phalanx. the -The end pa"s addueton digiti II is Pre_sent in all of the gr-ouPl and is siinilarly developei in all. It has a small proximal belly a_191g thin tendon extending to lying in the flexor groove -ani the insertion. The iize of the belly varies but without apparent significance. The exact position of the belly and its ielationsf,ips to the fleror haLlueis breuis or abduetor di'giti IV appear to ba a functional variable. Sushkin (1905:54) said that ohiy the nost distal part of the tendon of this muscle was present in Mi,Luusbut my-specinen hacl this rnuscle typicalfy 4e- veloped (as in Buteo or AquiLa). Nitzsch (1.866:L56) noted that this nuscle is nissing in Gypa7tua. FLeso? Hallucis Bteois The fLeror haLLucis brepis is single or two-parted in the accipitrid and falconid; the two parts insert separately when present. In these groups, the origin is proximal, more so in ttre latter than the former. Usually, in the accipitrid, there are two tendons and the bellies are fused (separate only in Tongos). The tendon of the pay,? erternus- of th-e lypicat acci- pitiid inserts broadly across the base of the phalanx' the FALCONIFORMES 73(27 3)

tendon fanning out somewhat just before inserting. The tendon of the pare internue is somewhat lighter in weight and inserts narrowly on the medial side of the base of the digit. In Elanue and Gampsonya,the pans internue has the wider tendon whereas in Chondrohieras and Avi,eeda the tendons are of about equaL width. Pandion is like Neerosyrtes in having only the pd?s internue present. In Ealeo, renaeidea" or PoLihieran, onl.y the pa?s esternus is present whereas in the other falconids both parts are well developed and the bel.lies are entirely separate poLyborus) or fused proximally. In the cathartid, this muscle arises along nearly the entire length of the second netatarsal and the belly is wider both proxirnally and distaLly. The insertion usually covers the proxi- naL end of the hallux phalanx. and, in some, the adjacent connective tissue. The proximal part of the belly gives rise to a tendon which inserts 1aterally. More distal fasciculae insert on this tendon throughout most of its length, but the nost distal ones insert upon the medial tubercle of the base of the phalanx and even across the ventral aspect of the deep flexor tendon on the lateral. tubercle as a perforate insertion. Perforation of this insertion occurs in some accipitrids. In Chondrohi.erax, both tendons of this muscLe enclose the deep tendon at their insertion, the lateral one inside the mediaL one. Sagittarius has only the pare internue, the belly is Iimited to the proxinal half. Lumbni.eales The Lunbrieales is best developed in Pandion, but it is present in all of the falconiforms. Cathartes differs fronr Coragyps in that this sheet inserts directly on the pulley of trochlea III, secondarily on the pulley of IV, and partially on the pulley of II. In Coragype the main insertion is on pulley IV, a weaker one on pulley III--and no apparent insertion on the pulley of II as in ihe other falconifoims. This muscle is thin in Coragyps and fairly thick in Cathantes, it is smallest and bandlike in Vultur and Sareoramphus. In Henpetotheree, only the slip to IV is well deveLoped, that of III vestigial. Pandion is unique in that there is in insertion on the posterior, lateral wing of trochlea IV as well as on pulleYs IV and III. Sunmary and Conclusions The interpretation of myological variations is contingent_ upon conclusions regarding their value. Such conclusions can be r-eached only after a consideration of the opinio_ns of the various contributors to the subject and a resynthesis of our knowledge. Garrod's wide (but unsystenatic) experience with the com- in his articles of L873 parative-and anatomy of birds is summarized L874. He considered the thigh musculature sinplest for analysis because of the variation in the presence or absence of a grbup of five muscles. He used a letter to represe_nt each of four muscles and a plus or minus to represent the fifth (the reetus fenorie of this account, Garrod's ambi.ens), and he thus ciraracterized the thigh of each species by means of a formula. 74(274) M. JOLLIE

On the basis of the presence (at least in some membersof a group) or absence (in all nembers of a group) of the reetus femonie, he divided birds into two nain groups. Beyond this he utilized variations of fornula as indicative of relationship or lack of relationship within groups. He pointed out (L874:LLZ- 113) that on the basis of his formula species show no individual variation in their nyology (ignoring obvious anomalies and considering only the presence or absence of the formula nuscles); that genera show no variation among their species, or only very slight variation involving a minor variation of forrnula, except where species have been lunped into unnatural genera; that fanilies show no, or only minor, variation in the myology of their genera, variation which involves only one element of the typical formula. He considered that any greater variations than these require further separation in a natural arrangement. In his view (p. 114), 'rAn inspection of Plate XVII is suff icient to show that the formula of a bird is not of direct value in estimating ordinal characters. . . " but by neans of his fornula and other anatomical characters (caeca and oil gland tuft) the known groups of birds were rearranged. Fi.irbringer (1888) laid stress not only on presence or absence of a nuscle but on its form, relative developnent, and relationships to others. His systen lacked any means of simple gxPrel- ion, like thatof Garrodfs, and because of this can be utilized only after exhaustive comparisons. Fiirbringer considered the myoiogy of the shoulder and breast most valuable and useful for the determination of relationships. Beddard (1S98:77) listed l-7 muscles which show significant variation in birds, nost of which are in the thigh. He pointed out (p. 76) that, "The muscular systen of birds is renarkably constant for the species, the number of variations being apparently, comparatively speaking, but slight. . .The muscular given or even system too is apt to be very constant for a -genus a- larger divisibn." Like his predecessors, Beddard made little real use of muscle differences. Hudson (1957) reinvestigated the thigh musculature of the bird and extended the fornula of Garrod so as to include other muscles and structural features of the hind limb, which show variations. His work indicated the possibilities for success offered by careful, comparative studies. He corrected many errors of Garrod and indicated that the taxonomic value of rnyology is largely "yet to be determined." As a first attenpt at utilization of hind linb musculature in deternining the relationships within an order and in differentiating orders, Hudson (1948) gave an analysis of the falconiforms. l{is conclusions are similar to those of Garrod (1874:1.13-114), differing only in that he separated Pandion as a suborder distinct from that which contains the Accipitridae and Falconidae. Fisher and Goodman(1955:L23-L24) cast some doubt on the value of Garrodrs formula and Hudsonfs nodification of it, at least in the cranes, where both individual and species variation in the presence and development of the two parts of the eaudilio- FALCONTFORMES 75(27 5) femonalie occurs. Historically speaking, ro one has as yet successfully es- tablished taxonomic value for myologl although many have expressed faith in it possibilities. Myological characters have so far been of purely secondary inportance. From my own dissections it is apparent that:total sinilarity of nyology (based upon presence and absence of nuscles, shapes, sizes, and interrelationships depends upon the closeness of relationship, general body proportions, and sinilarity of life habit (behavior). It is to be assumed that as divergence of species in tine takes place the differences in nyology, as of any norphoLogical character, tend to increase. ltlhere convergence has occurred the superficial similarities should be betrayed by differences in basic pattern. Generally speaking rnuscular structure is as conservative as (but no more than) bone structure since the two are interrelated. Because evolution is not a predictable process, two families of a group may be closely similar in the structure of the limb, and a third nay be different. The actual tirne of divergence between the three night be the same, but on the basis of myology it would appear otherwise. The absence of a particular nuscle, or other minor changes, does not necessarily indicate a lack of relationship or even Lack of close relationship (aee Fisher and Goodnan, 1955: 123-t24) . Fiirbringer (1888:1055) cited a hypothetical instance where myology could be utilized to denonstrate relationship when the underlying bones differed strikingly. I doubt if such a situation actually exists. Here we encounter the difficulties of interpretation. One nust define the basis upon which a similarity or dissimilarity is said to exist. It is possible that the osteology of two species night be thought more dissinilar than their nyology, and yet when approached by a different set of criteria, the reverse might seem to be the case (i.e. problem of Fregata- Cdthaptee relationship) . In terns of rnyoLogy, conservation and adaptation are well represented. The distal wing musculature is conservative, since most birds f1-y and util.ize relatively similar notions. These movementsvary in rate and style of flapping as well as the pro- pensities for soaring or sailing. Variations in the distal parts of the wing (not so much the presence or absence of muscLes but their relative size, form of belly, rnodifications of origin and insertion, and general patagial pattern) do occur and can be correlated with the shape and use of the wing. Two species , even though closely related, will differ in detail if different proportions have been developed. Less conservative and more indicative of basic lines of specialization is the shoulder and breast musculature. Within the falconiforns there are the basically different patterns of the cathartid and aegypiin. Both are adapted for soaring and in general there is a sinilarity in body proportions and in be- havioristic adaptations of feeding (the same situation described by Steinbacher, 1935, for the foot). It is interesting to note in this regard the detailed anatomical similarity of the raven 76(27 6) M. JOLLIE and crow but the difference-in their flight habfts*- These-ca5es iiidiEiiE ifiit-t[e-sime-mii5cle--st-tern can-be used effectively in different ways and different nuscle plans can be adapted to the same type of'movements. Agreemer-Itin detail is not conplete--for examplb'the angle at which the wings- normally are held may.differ. 'The hind finU musculature probabLy shows nore adaptational nodification than any other region. This lirnb nay be- scarcely used (rroehili.dae, Fnegati.dae,_ capni,muLgi.dae-), pgYerfully de- for grasping piey (hawks'_9t1t)-or for clinbing (parrots)' "uiop"aeionlatea foi waiinE ltreions)r well develope4_foT.running and walklng (galliforns), modified for swinning (Gaviiformes, Anieriioites) or for-perching and foraging on the ground and in trees (passerines) Tia) trunk and vertebral coLumnmuscles will show some varl- ation because detailed movements denand specific nodification in region. Boas (1.929) described differences in the cervical ""inirscufature and corretitea these with the length of the neck and tnJ-feeaing novements of the head. Sirnilarly tL" muscles of the jaws show ilodifications following those of the forn of the bill ind its use. These nodifications of trunk and head, however, are generallye of a detaiLed nature. The'differences observed in the falconiforms cannot be ascribed to adaptive radiation, singe the-comnon habits of this a ;;;;t-rhoutd froduce paralle| similarities. There is as wide ;;;ta of habi'ts arnongthe accipitrids as within the falconids, there is basic agreement in rnyological detail within ""a"y"tgio"irt but not between gioups. Thus, the evidence presented by [tt" iat coniforms suggests that the several myoLogical patterns in the first radiation of the avian type observed were developed-secondarily and that these have becone adapted for a predatory or vulturine existence with little or no actual convergence. Among the falconiforms over 50 muscles, or_their parts, are observed io show usable variation. In Tables 19-20 various comparisons are made (sinilar to those of Gadowor Fiirbringel). th; groups of the ialconiforms are compared one with another and *ith"to.ri genera of other orders. Each grgup is--represented by a symbol, ised when the nuscle is distinctive. When two or nore grolpt have comparable mus_cles, the symbol of one of the groups is u'sed for all- (A in preference to the other symbols among falconiforns). The preparation of such tables presents the problen of ascer- taining wiretirer the condition of a muscle in one grgup {"tls into the exfected limits of variation of another or whether it is sufficiently distinct to be regarded as uniqu-e. Th" decision is subjective. an arUitrary one and these comparisoJrs arer -therefore, I have tried to be conservative and have indicated some characters as shared that other workers night feel are distinctive. The use of such a table of-conparisons is nisleading in that the relative value of a detail may vary frorn insignificant to important and yet constitute a unit character in either case. closely foi exanple,'GaLLus iir Table 20, the cathartid seems to be more related lo than to Fregata, but the cathartid and GaLLus exclusively share only 4eoof features while the cathartid and FALCONIFORMES T7(277 )

r{ €tl o o o 03 F{ t{EOG d tr \o.o ta !- ao o r{ E 5 !- E.'-{ d }1r{ (\ O {J O A?l U..i F€ rd F- nddo (, r-- otr!n+r.ll E o c! o(, t{UotrA o ,E bo5 q{ O +r.H +r cd' .F{.C O }r d +r H o€ cd snAJoJ 4 |4rri<'. C .r'l +, t-{ U .r{ 3 c6 r-ia O F{.r'{ .F{ H.d O OF{ k cdd r{ dbO E .F{}>Ucdd .-?ln OO E fix O .r{ l|tP u).e E,-c a . +J u d tn o€{.)+r .l-, l.{ Flr{ U U t{ g tr U H O'r{ cd i: cd d Gt O.F{ .r{ O O d C o o d tr rD.C b0 F{1.rtr t{1.1t/').F{O € >U .F{ .r{ .F{ .F{ k +J d t-.'-{ P+J.F{t{HO Oe E f{ta tJt<+Jo ao ,o'9 gon O F{ d.r{ 5 o O k 5 5 i') O.'r d.r{ $t o tnq{ b0 ll a ..lhl .'{tC€cd'dhEd cdOoH.'-{€tn r{F.{ a?t .'-{-d.Fll*{ H(ll > d'- u o t{ tr.F{ c o.F{€ 'd l^'o h H.o t{ 5 d 5 5 F.. +,r €.,..t O h d t/t u€ o (rt.C 5 AX 5 5 o.rr o.r{ (! o uFl C tr (!l5o.r1pr.:!fo .r{5 d o "d E ^O c6d dooPE "d hF{ F i trt <'dU ElndOt4 HooEu Ut{UHOt,l >p.hoq) d.F{ tr .rt ll ta..tx (aolnS pcd..{"o5u) (6(do+)+r . .F,l +J 14 tr O OrCr+.{.F{ t{ (^ 0n t{ O la.O t{,.o tn 5 (/)r-{ U H X tr Or.q O . k Or-{'{ O Od'6 h t{ 5 O (l.'.{lJ O ta (u k E O U o.r.l o.Fl r-{ +Jr{ U/| k cl . (/t bou tr d O O O O o UF{ U t O.F.rd E .r{ U k d,C O |noO'..Coot/|>ou|cldtnO AUl >(/).F{t^(^ qa Olln(6r-{tr.O op tr o.dtt H l/l d O O k t{ tr cl5 o c') O t{ F{ (/t lr F{ F{ 5 O.r{ d 5 t{ (0 t{+r bOU U (d d d lrF-{ O d tr.F{+r C.F{ P 6 d O rd d .O O E OEUFE At{ O ln.F{ d d Ad.P tr1 Up AOd U (d€'-{ gAh p.A Cl tr Q.r{ ll ll OO+r f{hkXkSO d +r.oOhdO O) F< o d (r' o(JFa 'duu) o.|J+r(.)+ruEl F{ ooti+rFr$ .P 78(278) M. JOLLIE

vvuv

tr{AFA

t{ t{ flrf&<(9A(J< L)(J

t{h h S{ trrtl{CJCJAF{< FiCJ (J

{{C.lrrifrr<<

14 H Ff Fl o F? AA

E aa<4aa<<<

th P'FloP ,.c Ed.'-rdu|E .J,o'r,|dJ-6taFFllnF- ootro uop Ho''ru(no .r{.r{t.Ft.'.{.r{ 6 b X F 1r g t q +'5-O gH 5 .t.r q{ O F{ b05 H 5 U r'. 1. h 1.'q6' .'{'-r.t{ -HF.{.r{F{oF{p.r1 d otuut x qoqo.;6+lt oidcjE.;o. '-ro'o> r{ rn u u g(6'd o O o ul F{ t{ P Q.pAE tr{k e.I - - !dr, t ddF{OF{ FTFtOtrOOX o l.Oj-O.;6J;.;ArO..l'-' Oh€,o rNt+r'o'.{gtr.c) D-- ii-.tnO.o+r.r{bop.tr>p rnbocdo uboAtn+i FALCONIFORMES 79(279)

'o q) . . .F-. {3 o,c o xoo-f€Gror+l€l +) +r +) C) C\ F\ \O rJ'ltn F t Ccdt{ - -El ot+{ u E A A 'L E o.x o Q{J.FI O €t{5d ,r ooci{-{ootn o\ h E.r{ t{ rN (B D- L boh o b0 C\ (0ooq lN0) .Fr. 'L tn{JhO..qO|{ trl\oe{@€(\rOOFfle\t o\ .HCdUCd r- \O rA at n *O.OlHl r.- ,t €C o(0c) .'''l O O O bOO 'lc t{ > t{ ct r.{ O"OOh6O+rO v!4* o.o d +r tr3 +r(aRltn oE l--,o o 'l' Cq .r{ou>. o o€'d 5 l{ o .c;; *; Jl*l ,; E t{ H do r+ ra ";F- O Atda.r{ A0) - - -q.l L ox dH t t{ O+r t{ 5 OF ,r b0 go .4 (! ln botr o +r UU* .F{.'{ O O t+l .^| f.. b0 k .ll> 'L oo Srdx t{ . AOt'.+r(D+) tJ{ H t{ Otr + -oHcd r{5dUcU .r{ O Fod{J} *p'p OO .l-r.r{ + tPc6Otn>\ - - O.t{ > r+{POOrJFi..O u);lJl- .-:J .F{ b0 O l. O Pl| CO -; ^; k .Fl t{ € X.r{ X 6 €'o * X tr tD r+{ r-{ 1.. r-{ -R.l +) k (d.q o F{ o c) H tr ut{c d +r (IlP> ct O.F{ E O H rd.,r r.{ +J U{. E ot+.r- b00 E ln AU Soo(6bo , +r (O oo 5".{f' oD A OFr bp-ooo .Fl E ti Xo tr'-lE O ()t .t . oP drc .+JbooQ+rX lorlF{ t{ O tn (dt o (l' U OOI ,-{-t tnL c{C+rh tr F{ ooh tD d O "Fr€ '{.1 k k Or. OOOAt..qOq) o(l'x r{}U OPl

Fnegata share L08. 0n the basis of features I judge to be most and Fregata much closer to each sigiificant, the cathartid -are otf,er than either is to GaLLus, but such a rel"ationship is not indicated by the table. Fron the tables the following conclusions are suggested: 1. The different groups of the falconiforms, including Pandi.on as a separate entity, show . about the same number of differentiating features as do species of other orders. No close interrelationships are indicated. Z. The cathartid and sagittariid types stand out as distinctly different from each other and the other groups (or genera). 3. The accipitrid (with or without Pandion) has few unique features and exclusively shares none with the other groups with which it is usuallY Placed. 4. The suborder Falcones compounded of the accipitrid, pandionid, and' falconid types, cannot be defined on the basis of sharing nyological peculiarities any more than it i-s definable osteologically or by ptilosis. The first of these items is actually put to the test in a discussion of the others. The second implies that we are not e;;iing with "prinitive Accipitre-s;" i! suggests independent diverg6nce. IiI contrast, th; cathartid sug-g-estively agr-ees-with which is f""i"|", There appears to be ?-positive affinity here ;;t"i.ttdic"t"a atno^n-gthe falconifbrns . Sagi'ttari'us- shows evident specializatioi but not from an acciPitrid or falconid tlpe. Nor dois it seem likely that the falconid or accipit_rid could stem from the sagittariid type. A conmon ancestor for all could not be viewed fegitimately as a memberof any of these groups' As to the tf,ird iten'it should be explain^ed that separation of pandion from the accipitrid does not esentially alter measurement of the degree of di3sinilarity 9I similarity of the latter. itre-vafue of sEparation lies in the idea that the extreme of a gio.rp can show i degree of independence sinilar to that of an unrelated sPecies. The accipitrid shows few unique features and, therefore' can hardly share lnany with Pandion. One night say that most of the features which are plastic have been altered in Pandion; those that are not, are shared in a variety of ways with the accipitrid and other groups. The distlnctiveness of Pandi.on is somewhat reduced by the variation from the basic accipitrid type observed in the pernin and elanin kites. These do not bridge the gap; rather they denonstrate the plastic nature of the aberrant structures. The comparison between the owl and accipitrid is much the same in terms of percentage of dissirnilarity, but this is not the whole story. The differences are more significant; further, some of the special features distinguishing the owl (particularly FALCONIFORMES Br(281)

in derinal musculature) are not among those evaluated. The assuned commonancestry of the accipitrid and faLconid finds no support in myoLogy. The hind linb musculature, which should show some evidence of convergence, beyond the increased size of the shank muscles (for grasping), inste.ad fal1s into two clear-cut styles, which do not overlap. Nor can one be derived fron the other. The style of the falconid can be considered the more prinitive (having a eoLeus and medial eemi,membnanosue);it cannot be any more successfully compared with the sagittari|id or the cathaftid type than with the accipitrid. MyologicaLLy the cathartids can be split into two subfamilies, the first containing Cathartee and Conagype and the second the larger condors. This division is founded on the presence or absence of the eaudiliofenoralie and the presence of a single or double tendon for the obturator esternue. The nyological constancy of this group (many ninor variations exist) can act as a nodeL for the amount of variation to be expected in the subfanily and fanily. If anything, the range of variation is less than that expected in a faniLy (Fisher' 1946:709-710). Someof the present subfamilies of the Accipitridae are not represented in the dissection naterial, but pertinent facts can be- added to those already known. Pandion is usually considered to be most closely related to the pernins, but comparisons of musculature indicate little essential agreenent. Points of similarity are few but include close agreement on the muscles of the abdoninal wall and the heavier whiter tendons of the eatensor digitorum Longus. The tendons of pandion mark the extreme, whereas- Condrohieras and Auieeda are halfway toward the usual accipitrid condition. Elanus has a shank that,superficially resenbles itrat of Pandion. The pernins appear to differ by the absence of the reetus femoris (ambiens) , less modified abdominal nuscles, and more diStinct bellies for the superficial flexors of the shank. ELanus, including Gampsonyt (and Cheltctinia), stands out as a peculiar genus by virtue of the insertion of the reetus femoris into the lateral part of the superficial flexor of the shank; the forn of the bellies of the superficial flexor; the eaudilio- femoralie arising by a narrow tendon from the ventral, anterior point of the pygostyle (appears to arise more forward than in others); the veiy weak pe"onaeue Longua; and the fairly thick and white estenson digilorum Longue tendon, which resembles those of Pandi,on, Chondrohierax, and Aoieeda. The shank rnusculature resenbles Pandion in its molding, indicating a powerful leg and foot in this snall hawk. The aegypiins (only Necrosyrtes and lotgoe exanined) have a peculiar eaudiliofemonalie, which resembles that of Pandi.on in its long insertion tendon. The dernal muscles of this group are particularly well developed. Neophnon Garrod, L873:635, fig. 3) differs from Neerosyrtee in the fortn of the eaudiliofemorali'e, resembling the typical accipitrid. The rnyology of such genera as GAps and Gypaetue (Nitzsch, 1863; 1866) appears to differ 1ittle from that of the typical accipitrid. Myologically the aegypiins are not supported as a subfamily. 82(282) M. JOLLIE

Aquila (uroa'dtue), Buteo" Cireue, Aeei'piten, and fctinia show little difference as conpared one with another. A myological description of one suffices for all. Gymnogenys,which is well narked osteologically, differs little (Beddard, 1889) fron this central array of accipitrid species. Although there is rnarked uniforrnity among the f alconids, the Falconinae and PoLihieracinae appear to differ in that the pars internue of the fleaor hallueis breuis is lacking, the peronaeug Longue is small and does not overlap the tib'i.alis anterion (a feature shared with Henpetotheyee as a specialization), ald thg nerve of the semitendinbeue passes lateral to the pa?s nedius of the biceps fenorie et eemi.membranoeuarather than through the belly. bolihieras is peculiar in the rectus tendon passing nedial to the insertion of the iLiofibularie, rather tharr lateral, and in the resultant reorganization of the superficiaL flexor muscle origin. Furthermore this genus has a small slip of the soleus fused with the gaatnoenemiue. Both of these features represent advances over the prinitive style, a change which is in agreement with the specialization implied by the reduced body size in this grouP Herpetotheree and l,tieraetun resenble the Polyborinae in the reductioir of the pars peetoralie of the cutaneous maxinus and the (fro-n in developnent of th-e thrbat nusculature; they,differ -a11) that the eoleus is partly fused with the nedial part of the gaetrocnemiua. These two genera differ rnarkedly in details of iheir tarsornetatarsal nusculature and in the developnent of the -peronaeue Longus, The Polyborinae are weLl narked by the style of the tensor patagi,i, bneoie tendbn, which resenbles that of the accipitrid in ils division into two parts. This is possibly a retention of a more primitive style in contrast to the unbranched condition observed in other falconids. VISCEML ANATOIvIY

Introduction Four groups of characters are considered here; these are the convolutions ol the gut, the presence of both ovaries and their relative developnent, the nain arteries in the region of the heart, and the syrinx. The first and third of these are investigated most easily in fresh specinens and with more difficulty in preserved naterials.

Comparative Anatony Gut Gadow (1,879, 1889; Gadowand Selenka 1891-) described the convolutions of the gut and a system for their analysis. He described various loops and-classified the relationships of the arms of the first four of these loops into several categories. He (1889:309) intes- pointed-tina1 out that the accipitrids tend to have five or six loops and these are- orthocoelous (arns relatively straight) with a tendency to teleogyrous (ends of loops curled) or mesogyrous (irregular spiraling of loop). FALCONIFORMES B3(283)

The convolutions of the gut werb studied in a different manner by Mitchell (1896, 1901). He pinned out the gut on a flat surface and recorded its relationships to the mesentery and to the hepatic portal vein and its branches. Mitchellrs systen has considirable-rnerit and was used as recently as 1938 by A1- Hussaini in his study of. the Aegyptian kite. In order to compare the guts of the different falconiforms certain introductory concepts are of value (fig. L97). These involve both the basic plan and the finer details of the gut. In its sirnpl"est form the gut is a straight tube from stomach to cloaca.- Anong most birds two segnents of the sinple gut have retained theii original positions and these are the stomach and the large intestine. The falconiforrnrs stomach has increased in length and its left waL1 has buLged out Posteriorly-, leaving. the pylorus at a point on its. right, dorsaL aspect. The stonach is d-ifferentiated only poorly into an anterior proventriculus and a posterior, sacl.ike ventriculus. The tomach walls are disten3ible and in preserved specinens the shape of this organ varies greatLy. Gadowrs (Gadowand Selenka 1891:641) comments on the forn of the stonach are useless' if one has only' preserved naterials. ttre-omparbtive Large intestine is relatively straight although-it-nay show minor klnks which correlate with its distention by food. It is proportionally short but with the shortenilg of the body cavity- of- the bird ias cornparedwith the long bodies of prini- . tive vertebrates) its anterior end is separated frorn the pyloris by' less than its length. The snall intestine is the part of the gut in which elongation has taken place. The snall intestine is usually thrown into four najor-1oops: the duodenal, meckelian-, supraduodenal and precolonic (supiacaecal of A1-Hussaini). These loops can be identified by convenient reference points. The duodenal loop is a typical "closed" loop, that itr, tlte descending limb (section passing posteriorly) parallels and lies close to ltre ascending limb. It arises at the pyloris and passes posteriorly a1-ongthe-dorsal, right aspect of the ventriculus. it extends'around the posterior curve of the stomach and then projects into the body space or is thrown into a small coil Iteiogyrous state of Gadow). When coiled the distal end curls iorwaii and dors aLLy in an anti-clockwise fashion as viewed ventrally. The ascending linb lies dorsaLLy and,follows the course' of th'e descending onel at the level of the pyloris it passes dorsally and nedially, close to the dorsal body wall. frilm this anterior, "body wal1r" position the neckelian loop begins and the greater part of the leqgth of the gyt is involved in-this sectionl This loop at its midpoint, broadly speak-ing, which appears a small out- nay show a "vitelline" diverticulum -as potketing or merely as a thickelingr-of sqotr or- the wall of the Lut. This diverticulun marks the point of attachment of_the- gut io the yolk sac. This diverticululn can best be seen in fresh nateriai and is, in fact, difficult to locate in preserved specimens unless it actually projects ftqT the gut wall. This 16op is of the "open,'type and its ascending linb ends on or neai the dorsal midpoint of the ventriculus and nay' or may not' 84(284) M. JOLLIE

PRECOLONIC

MECKELIAN

MECKELIAN 3 SUPRADUO- DENAL SUPRADUODENAL

DUODENAL

Fig. L97. Two styles of illustrations of the convolutions of the intestine A. and C, Elanus eaeruleus; B and D, Conagyps atratus; E. Pandion haliaetus. FALCONIFORMES B5(285) be bound to the stonach wall. The neckelian division may be a' synnetrically disposed or spiralled (nesogyrogs, as opposed to rbguLar spirals, cyclocoelous) with the nidpoint lying dorsally and centrally. This Loop is essentially undeveloped in Elanue. The supraduodenal loop is a closed loop arising in the region of the dorsal wal1 of the ventriculus; it:may be bound to the stomach wall. Its descending linb passes posteriorly, later- alLy (to the left) and ventralLy to forn a half circle, its distal end being directed in an anti-clockwise coil as viewed ventrally or projecting dorsaLLy and anteriorLy. The distal end of this Loop nay pass between the stomach and the end of the duodenal loop (faleo, Polyborue, Conagyps). The ascending linb passes dorsal to the descending and is always bound to the stomach wall for a short distance. The meckelian and supraduodenal loops in their simplest state encircle an area of mesentery in which branches of the hepatic portal vein 1ie. These loops usually have-secondary nay be thrown into snalL tertiary ones. lobps,- which in turn Fron this point of attachment to the stomach, the gut- precolonic passes to the laft and dors aLLy as the looP. - !his. -portion generally has irregular folds and kinks and usually there is a single, snall open loop. This loop nay be essentially lacking \tlanus) or there nby be a conpounded one (Cireue). The section ends at the beginning of the large intestine. The large intestine at the point of transition usually has a pair of lateral diverticulq of varying size. Diverticula are tatking in the'cathartids and there is nq _easily identifiable point where the small intestine leaves off and the large intestine begins. Other-features of the gut are worthy of comnent at this point. liver is bilobed, the two parts about equal -size The -being -of and disposed symnetrically to either side and below the proventriculirs. Th'e right lobe is attached to the dorsal body projection which encloses the post iuaff by-A a posterior nedlal cava. lirge gall bLadder is present and there are two bile ducts, one fion-the gall bladder and a second fron the region of anastomosis of the two liver lobes. The bile ducts enter the smaLl intestine in the region of the distal end of the duodenal loop. The pattern of the biLe and pancreatic ducts probably tould not be used for the characterization of groupl. ' Thl) spleen is poorly to well developed; it is located in the dorsal-mesdntery of the stomach, anterior to the curve of that part of the gut attached to the stomach wa1l between the supraduodenal and precolonic looPs: - The tongue is- well developed in all of the falconiforms. The cathartids have the dorsal lateral margin of the tongue denticuLate (fig. 198). The falconiforms all have expanded the esophagus inio i crop. The crop is highly distensible and cases are offen cited of clrrion-feeding types being unable to ft.y (not Corag.ype) is because of their gorged condition. Cathartes - peculiar in having a-dense, hairlike, mass at the pyloric exit of the stomach. In comparing the various groups of falconiforms, there are no great differences between then. The accipitrids show the 86(286) M. JOLLIE

ffi{}1 s//t A B

Fig. 198. Dorsal views of tongue and Lower jaw of A. Conagype atiatue, B. Buteo ianai'cenei.a. greatest diversity as gxpgcted in view of the number of species. Subfamil.ies could-not be differentiated with the naterials at hand. The' sinplest gut is that of Elanus L97)- in which the duo- denaL loop cin almost be considered op9n. The meckelian Loop is scarceiy more than sinple. The region between the rneckelian and supradirodenal Loop is- sLightLy open- (Like- the duodenal) . A precolbnic Loop is ab-sent, the gut showing only a slight propensity for kinking ' distant from Elanue but has typical Aui,eeda is noI too ' closed duodenaL and supraduodenal l-oops. The meckelian- loop is Uuttut deveLoped and i; asynmetrically disposed in the body cavity . Aeci.pitet etriatu-s is as sirnple as Elanue with the exception thai a distinct precolonic loop is present and the firs't and third loops are Llosed. The extreme of cornplexity is - found in Cireus in irrhictr the neckelian loop is spirally disposed, approaching the form found in Faleo or in the cathartid. The siiraf arringenent of the neckelian loop probably vSries in its plrfection ii aifferent individuals. In general, meckeLian -the ioop- of the accipitrids is asynmetrically disposed.. ' affects the neckelian The length bf tne -Cineusgut, which largely section, is iariable. has the longest gut in proportion to its 6ody size. The guts of Ior,goe' Neeroayyle_9' and naliaee- tua are a13o long (see ionnents of- Gadow1891:641). Tl-t" length of the gut is coirelated with the anount-of looping and, usuallyl the propensity for'mesogyrous disposition. pand,ion is iistinttive in that a recognizable supraduodenal loop is ab'sent and the enormously eLongated rneckelian section is FALCONIFORMES BT(287 ) dividod into a large series o'f closed, straight, secondary loops (20-?,5 in nunber). Haliaeetue has a somewhat sinilarly elongated rieckel.ien section but its secondary loops are not as closed and both the supraduodenal and precolonic loops are strongly developed. In Haliaeetue, elongation has also involved the duodenal loop which shows secondary folding whereas in Pandion elongation has- involved nainly the neckelian section, Pandion also shows a proportionally longer large intestine anq a Ygry 1?tgu,.kidney- snapea spleen. Gadowfs (Gadowand Sel.enka 1891:642-644) com- nents on Pandion appear to be based upon a single specinen in which the disarrangement of the digestive track led him to record severaL points not substantiated by my specinens, nor to be expected in such a bird. - The falconids are quite like the accipitrids. FaLeo has the is irregular- neckelian loop- coiled, whereas the gut of Polyborus ly disposed. In both ?aLeo and the cathartid the base of the m'eckelian loop is closel-y appressed and bound by loose connective tissue tb the dorsal wall of the stomach. This does not seem to be the case in the accipitrid. The accipitrid and falconid are generalLy sinilar but appear to differ in the developnent of the sPleen. Tire cathartid differs in the forn of the tongue, a fanily criterion at best; it also differs in the consistent and fairll synmetrical coiling of the meckelian section and the absence of the colonic caeca.- The relationships of spleen, liver and stomach are somewhat at variance with the accipitrid picture (such as the usual appearance of a senidistinct pyl-oric section io the stonach in the acciPitrid). 0n the bas is of the gut there are three poorly defined groups: accipitrid-faLconid ,, Pqndi,^on,and cathartid (Sagi.t- iar.iia not sebn) . The gut of the falconiforms shows a much wider range of variation than that of the tubinares, although the lattei have a somewhatmore consistent diet than the, falconiforns, according to some.* Gonads (1940) Rand (I942) hlve stanley and llritschi and -suppLied much of the- information on bilateral ovaries in the falconiforms, (Table- 2L). Apparently the cathartids never show a right o_Yaryand thus mark a-rnore modified stage than other groups . Pandi.on also has no right ovary. The kites tend to lack a right ovary and the largelt, most speciaLized forms such as the vultures are sini- larl In such sfecies as Buteo iamaieensi,e or AquiLa ehry-ea?-tos the right ovary rnay be lacking or_present. Aeeipiter and Cireus agree In the strong developngn! 9f-b9tl ovaries and stand in *:* * t * * * t *I * * *-* * * * * * * * * * * * tt * * * * * * * Forbes (L881:17) notes that, "The Tubinares as a,group agree very closely together in the forrn of stonach and intes- tines posiessed by them, which have peculiarities not occurring in any-other groups of birds, and it is olly in the variations in foin and slructure of the tongue, in the nature of the armature of the mucous nenbrane of the nouth, and in the presence or absence of caeca that the various forms differ in any inpor- tant degree from each other." B8(288) M. JOTLIE

TabLe ZL. Occurrence and relative developnent of the right ovary in the falconiforns. Classes A Right ovary more than half the size of the left. B Right ovary one-fourth to one-half the size of the left. C Right ovary less than one fourth of the left but present. D No distinct right ovarial rudinent. Species No. exanined Class Cathartid Cathartes aura D Sagi ttariid Sagittarius serpentarius A Accipitrid Aviceda subristata 3 C.D Henicopernis longicauda 4 c-D Chondrohierax uncinatus t c Milvus migrans 6 C-D Haliastur indus 1 C Accipiter gentilis 6 A Accipiter novaehoLLandiae L0 A Accipiter poliocephalus L B Accipiter fasciatus 2 A Accipiter cooperii 16 A Accipi ter. ne lanochl amys z A Accipiter nisus 42 A-B Accipiter striatus 4 A-B Accipiter cirrhocephalus 3 -A Circus cyaneus 1_9 A Circus pygargus 3 A Circus aeruginosus 1 A Circus spilonotus z A Circus approxinans L A Buteo j anaicensis 23 C-D Buteo lineatus 6 C- Buteo buteo 7 C-D Buteo platypterus 4 C-D Parabuteo unicinctus 2 C Harpyopsis novaeguineae 2 C-D Hieraa6tus morphnoides z c-D Aquila chrysadtos 5 c-D Necrosyrtes monachus L D Pandion haliaetus z D Falconid Falco peregrinus 1_0 B-C Falco cenchroides 1 B Falco tinnunculus 2 B.C Falco sparverius 8 A-B-C Ieracidea berigora 2 B.D Herpetotheres cachinnans 1 C Milvago chinachina t c FALCONIFORIIIES B9(289)

and aquilin strong contrast to the buteonin -subgroups. itre right ovary in Aeeipi.ten has leen denonstrated as functional at LEast in bne case by Stieve (L924). He observed a specinen of Aceipiter gentilie in which two ovulated follicles wlre found in thir left-and one in the right ovary; the bird was sitting on three eggs. (The 9gg hag pa-ssed.from the right ovary down-the functionat left oviduct.) Th9-right oviduct is eithbr absent or extremeLy vestigiaL in all b.irds. stanley and witschi irg+o) pointed out that in the Acci- pitrinae thi' ovaries are nearly tlt" same size in the re-sting better it ase but with activity the left becones. larger ang is ieveloped. Their data'do not confirm this contention. Eirbryological studies (Stanley and Witschi, 1940) show that the nunbei of-gern cells in the splanchnopleule is equal on either side at early stages but as a result of secondary-nigra-. tion these cells become concentrated on the left side. The left side appears to have a greater attraqtive power fol these cells in Lat-ei stages and as i result produce-s a la-rger left ovary'--- sometirnes to-the totaL eliminati-on of the right. Stanley (l'937) stated that lack of development of the right,oY?rI is not due to failure of ahy one part to'develop but_to a deficiency 9{ il:.^. a,r.tott in th6 enbryonic gonad siie. stanley and witschi (1940) poi"i"a o"t that soire birds are unique-in that the right.gonad Ii-.o*posed onty of the medullary-portion. The rig!! oviduct air"pp!"tr by sbcondary atrophy bf- an originally well-developed enbryonic structure. 'These same workers stated that species with sma1l right ovaries tend to have a reduced right testis. The observations it ty study do not support this pienise; there is individual variition ianging froir'a right tbstis h-alf the size of the left to a right teitii considerably larger_than the left. Most males tend to have a slightly'1arg6r left gona-dregardless of the ovary condition. A tale'of CorqgAp? had the right testis about twb-thirds the size of the left-whereas in a specinen of Cathantes the left testis was just larger than the right' In Iorgoa, Buteo, Mi.Loue" DaPtni.ue, and Mi,craetun the two teStes *""6 rirughly equat. in a-single adult nale goshawk the left testis was just a bit larger. Arteries in Region of Heart The circulatory systen of the falconifoln !s- gener+TY. !lt9^^ same as that J"i.ti6ed'by Gadowand Selenka (1891), Beddard (1-898: (1873) 48-57) or Hynan (1942:35I-358, for the.pigeol). .Garro-d was tire firit to attempt a systenatic investigation of a paTt of of the-C""otids circulatory systen. ite described seveial -Patterns itt" and indicated their o'ccurrence in different groups oi-Uiiar. Th; ;uUject of carotids had olly- incidental consideration until Glenny-reopened discussion of the "main arteries i; the region of ine hlart" (Glenny, L955, for summaryof a series of-papers) with the view that their taxonomic importance had not been fullY evaluated. of the left fourth Glenny deiiriUua a ligamentous -vest!Se- (systernii irch), a liganenlous vestige of the right sixth arch ionnection to itre dorial aotta (ductus botalli or ductus 90(290) M. JOLLIE

arteriosus when a lumen is present and ligamentun botalli or arteriosus when the lunen is lacking), and liganentous vestiges of the dorsal conponents of the third (carotid) arch connected with the systenic arches (ductus caroticus on right side and ductus shawi on the left) which serve the dorsal aspect of the syrinx and esophagous. The arteries in the region of the heart are fairly sinilar in the falconiforms, (type A-L of Glenny, L955:562). Glenny (1941) described the arteries in the golden eagle and red-tailed hawk. He found a Ligamentun botalLi on the right side in both species. In most species, a recognizable right liganentun botatti coul,d not be found; it was (fig. 199) well developed in

e an o f (J gf E FJ I -o G Pf, o 90 o I o az F 68 o Eg g, E () = F -4. UJ !u lrJ gi F l^l o r XF o. at G Fg F (',\f I o G IJ *L- o z o = g . J o x z ,tr, T> to urfa

lrJ l- E o kz 63 J = x :< (J o 6 =() c! 2 o 3 z E o o

Fig. 199. Ventral view of heart and nain vessels of Pandion haliaetue.

a specinen of Haliaeetue LeueoeephaLus. Glennyrs description of the left fourth vestige did not apply to the specitnens I ex- aminedg there was a distinct connection of this vestige by a thread of tissue to the left pulrnonary artery at a point just outside the lung. This connection is the left ligamentum botalli. In the species I exanined, there were vestiges of the ligamentum caroticus on the left side. From these dissections, it is apparent that there are usually bilateral ducti botalli (or art-eliosis) and vestigial systemic connections for the nore anterior arches, at least on the left side. Among the falconiforms (fig. 199) the bilateral innoninate vessels each give rise to an anteriorly and dorsally directed commoncarotia and a laterally directed subclavian stem which FALCONIFONMES 9I(29l-) aLmost innediately divides to form several snaller vessels: anterolaterally and anteriorly dlrected axilLaty arteryr and a laterally directed pectoral artery which al.most imnediately forms several branches. The comrnoncarotid, a short distance fron the innominate, gives rise to arteries which parallel tltu course rof the vagus ierve, extend dorsomediaLly to the muscles of the neck and on sy_rinx esophago9s. The the right side only, rnedially to tfe -and region-of the carotid fron which these branches sp-ring lies doisal to the thyroid gland. Probably snall branches come off both conmon carotids to the esophagous but these could not be distinguished in the uninjected- specin-ens. The carotids of either"side pass nediall"y-and forward between the two bands of the longus c-ofti nusclesj the left carotid comes to lie superficial to the left The right systemic arch lies just dorsal to the base of the right iinoniirate and usualLy-does not show as viewed ven- tiallyl It curves dorsa1ly, rnedially, and posteriorly, passing j"it interior to. the right-pulnonary -It artery- at the point where that vessel enters the iigtrt lung. reaches the dorsal nid- it is line just at the 1eve1 of-the ?Pex of the heart and here joinei by the ligamentous vestige of-the left arch. Just io-terioi to thi; point, lt- givil off.a-ventrally directed toeliac artery froir its righi side aqd further posteriorly a nedian-arising and ventrally directed superior mesenteric artery. The accifitrids and falconids examined a1]. showed a well- developed, teit systemic vestiSg connected to the dorsal aorta; in^" cithartid's showed a nuch smaller one. The falconids ""adiffered in that they lacked a distinct anq separate artery iior itt" right carotid to the esophagus and,syrinx.. In Faleo there was such a vesseL but it apleaied to be a division of the branch (sane in the cathartids, Pandion" Necro- "cervical" slight--qnd eyil"{. On the bdsis of the carotids there were - di.fferences in the details of the pattern of the cathartids as ;;il;;;e-"ittr the others . Pandion was quite similar to the accipitrid. ^ Among the falconiforms, sufficient differences are not yet known to ilake a practical contribution to systenatics. Syrinx The syrinx of the falconifbrnq (fi-g. 200) has been studied by Beddard, and little can be added !o-his accounts of this itr,r.i,rt" (f Ag8--pp .475- 476, for accipitrids and f alconids ; p' q7g,-for pindion;' fot sagittari-us-i-pP: fgT b. 480' -4qL-!82, cathartids; also-1'903) . Alden Milter (1937:204- 205,) has dis- iussed refirence poinis for comparing the !)'rinx _of different ;t;ai"r. As in his account, th;) pesiulus is used here as the position. kiy, although it probably is not constant in_it-s Foit6"r (1882:52) issuned'that the position of the insertion of the intiinsic nuscles was constant and that the number of bron- ctti.t semi-rings could be deternined from the insertion. An exanination of-his figures shows insertions on the first or ;;a;;e, third, fourthl fifth and ninth bronchial seni-rings and aLso an asymmetric iniertion in the case of Pagod?omani'uea 92(292) M. JOLLIE

STERNOTRACHIAL

TRACHEOBRONCHIALIS

Fig. 200. Dorsal (right) and ventral (1eft) views of syrinyx of A. Cathartes auna, B. Pandion haliaetus , C. Elanus eaeruLeus, D. Polyborus eheriuay. FALCONIFORMES 93Qgt)

(figures 13-L4, p. 35). MilLer (1957) found irregul-arities of insertion involving both synnetry and the number of the bronchial seni- rings . The-sinplest syrinx among the falconiforns is that of the well-defined syringeal box; cathartid (fig.-snill 200A). There is no there is a pes3ular band which passes between the bronchi at their origin fron the trachea. In Catharted and Coragype-.two tracheaL rings are involved in the fornation of the "syrinx." _ UnLike all other falconiforms the medial postirior aspect of the bronchus fron the pessuLus distally is not menbranous (lacking the nedial. tynpani-form menbrane), al.though the semi-rings a\e thinner in thi; region. The dorsal aspect of the distal end of the trachea and the proxinal half of the bronchi are nenbranous' or at least present a gap in the rings on this surface. A bronchidesmui is wantiig. The syrini of the condors as described by Beddard (1905) agrees with these observations. : ' In generaLr'th; trachea and bronchi of the cathartids are nore nem6ranous-and less rigid than in the others and show the simpLest condition. The tnaeheobronehi,alie nuscles and lateral . The s ituation in and' rnedial tynpanif orm menbranes are l"acking - - the cathartib inarks either the beginning or thc end of the syrinx. It can be assumed that the prinitive bird had a syrinx, and that it has degeneiated in the clthartid. These birds are essentially voiceLess. The syrinx of the accipitrid is of the tracheobronchial type (fig. 2008; C). The trachel ends at the "voice box" which is iad6 up of-the last few fused or semi-fused rings of_the trachea. The brbnchi are separated by a narrow bar, the pe^ssulusr-which bears the semilunai menbrane on its interior surface. The first bronchial rings are incomplete rnedially, thus forrning a nedial tympaniforn n5mbrane on elch bronchus. T!t" tynpaniform rnembrane ii iriaest at the pessulus (the entire nedial aspect-9f the bronchus) and tapbrs gradually to a point jus! outside the ltttg. The second and tlird lemi-rings of the bronchi are usualLy in separate{ contact at their ends (ventrally and dorsally) and -.t their midpoints to form an ovoid membraner- -a lateral tyrnpaniform nembrane (the interannular membraneof Beddard) . The bronchi, near theii origin, are connected by a membranoussheet, !he- bronchidesmus 5f neddard, which extends frorn the level of the fourth or fifth bronchiai semi-ring to the level of the distal end of the rnedial tynpaniform membrane A11 the species Lxanined (Table L7) have a distinct syringeal box, although in Elanus it is weakly developed. Th9 second and thi;d broncf,ial seni-rings usually enclose a lateral tynpaniforn menbrane, but in Necrosyrtes there'is scarcely ? di:tinguishable gap. Th6 tracheobronchialis muscLes insert typically rlpon both. it b second and third bronchial rings . In some genera (Auteo and Elanus) the insertion uPon the third semi-rilg may be reduced to a menbrane or be lackin!. The medial tynpaniform membranesof Elanus are nearly in coitact at the anterior end of the bronchidesmus whereas in nost accipitrids and Pandion they are distinctly separated. The syrinx of Oryotriorehis (Beddard, L903, {ig. 19) is'most peculiar and divergent; it lacks conpletely the lateral 94(294) M. JOLL]E tynpaniform menbrane and has the intrinsic nuscles inserting on the fifth semi-ring Pandion (fig. 2008) agrees well_ with the pernin-or the typical accipitria. 0n the basis of the structure of the syrinx oirl could asiune that the call was like that of the accipitrid and such is the case o The falconid syrinx differs in that the first bronchial semi- ring is always fused into the syringeal box and the second clolely apprbssed to it; the lateral ,tynpaniforn-g.aq.is well developed a-ndovoid in shape, and the traeheobronchialis muscles inseit upon the second bionchial seni-ring or oeyond this on the middle ol the Lateral tympaniforn menbrane.* The syrinx of Sagi-ttani.us is briefly charactetized by B-ed-- (1898i480) 4sr strong box is formed-by tlt" last tracheal dard -and "A ring, the intrinsic fnuscles are attached to bronchial semi- rin[ 2." Fron the description thi: syrinx-appears to be as dislinctive as that of the faLconid or accipitrid. In lacking the syrinx, the cathartids contrast sharply with the other groups. The syrinx of each of the renaining groups-- accipitridl faiconid and (?) Sagittariid--is distinctive. Pand,Toncannot be distinguished-fron the accipitrid by the forn of this structure but the sternotracheal muscles inserts asynmetricalLy on the ventral q?pect_of the trachea (as in Canpophaga, gbddard 1898:510, _fi-g. 157) rather than late-rally. ' ttre-range of variation of the syringes within tlt" falconiforns appearl to exceed that found in the owls or tubinares. Within bich of the subdivisions, accipitrid, falconid, etc., variation is no greater and, if anythitgr less than that in the tubinares Air Sacs Al-Hussaini (1938) has described the air sacs of Mr',Luueand those of the cathirtid'were described by Bignon (1889). Gilbert a bibli- iiSSSl described this systen in the bird.and suppli.ed dgt"piry. Forbes (1881:32) described variation in the air sacs oi tire'tubinares wf,ict involved bilateral interclavicular air sacs in nost of the Procellariidae as compared with the fused iitgf" r". in peleeanoid,ee and the Oceanitidae. Watson (L883) dejEribed the unifornity of the air sacs of thq Spheniscidae interclavicular sa9): and Eddy.(L942) iii"gi" .Waterman lt"Y".* * * * * rt * * * * * * * * * * * * * * * * rt * * * *'* * * * tt * * Beddard (18982475-476) described tbe falconid flrinx thor- oughfir-i'Much'like the syrinx of Faleo is that of Hieraeidea beii.goia; t can, indeed, detect no differences. So too Tinnuneulue aLaLd.arius ind Engthnopue tesperti-nus. In Miluago ;h';;;;A; ind a. chima-ehima tire syiinx is- at first sight perfect- ii tirnif.r, Uut the intrinsic nuscles only,just.ge! beyond the Firrt bron;hial semi-ring Isecond bronchial semi-ring]. In Herpetotheree eachinnane:this divergence from the nornal falco- ;i;A syrinx is carried still furthei, the intrinsic muscle being attached to the first seni-ring "The syrinx of Polyborus |raeili,ensis is an exaggeration of the faLconine type. Th; first and second bronchial semi-rings are very proninbirt and wide apart, thus leaving a- very spacious r intrannular membrane, to which the intrinsic muscles are attached." FALCONTFORMES 95(295) made a comparative study of five domestic species

Sumnary and Conclusions The visceral anatomy of the falconiforms like the other anatomical aspects suggests that the falconiforms are composed of four distinct groups. Lack of relationship is shown by the sharing of various characters with other, unrelated, species of different orders. The study of visceral anatomy is as yet only poorly documented but deserves further investigation.

EXTERNALFEATURES

External features have been the main taxonomic interest of ornithologists. A sharply hooked bif1, relatively smal1 head in proportion to the body, relatively large size, strong wings and tai1, and usually, powerful, grasping feet with long curved claws have identified the falconiforms. Minor differences in the shape of the bill, shape of nostril, proportions of wing and tail length, type of scalation of tarsus, proportion of tarsus to wing, tail, middle toes, or length of claw are characters for the separation of genera and species. The choice of those to be used in each case has been left to the taxonomist. As a result no two groups are characterLzed by the same features and the individuality of each description is presumed to be as unique as the species or group described. The external features used to describe falconiforms are reviewed here only in a general way and without reference to the voluminous literature, which, to a certain extent, is summarized by Swann (L924-45) and Friedmann (1950). Although it is assumed that external features wi].1 continue to be useful at the subfamily or family leve1, it is not the purpose of this discussion to argue the relative merits of characters or to improve on existing descriptions at any taxononic level, below that of the family. This account is based on an examination of skins of all the species of the falconiforms with the exceptions of Leptodon fonbesi (probably an aberrant Leptodon palLiatus), Leueopternis poLionota, and Haematornis asturina.

Comparative Anatony Head The bill of the falconiforms shows a wide range of variation in proportions and yet there is basic uniformity within each group. That of the cathartid is most distinctive. It shows only a superficial agreement with that of the other groups in its hooked tip; it compares better with that of the Anhirnidae or some galliforms. The cathartid bill is long, relatively wide, and tends to be well rounded and inflated. It is peculiar in its high palatal arching, deep tomium, lack of a well-differentiated cere, wide-open nasal vestibule (which acts as an external naris), and the perforate nasal septum. This type is unique among birds in the open vestibule. Among the falconiforms it is almost matched by that of Neophron, Neerosyrtes, and GAps. The closest approach to the naked, soft-skinned cere is that of Neophnon, but 96(.296) M. JOLLIE in this genus the nostril and the inperforate septum are distinctly accipitrid. The bill and cere color of the cathartid varies but it is not yellow. The accipitrid bill is usually short and stout, the culrnen curves from the cere to the tip. The shape in the accipitrid can vary with specialization; it is weakly hooked an4 gallina- ceous iir appearance in Pernie and extrenely elongated and hooked in Roetnhanue (and in Heli,eolestee) . There are minor variations in the angLe between cere and culmen; the extremes, among the not obviously modified forms, are illustrated by Aeeipiter etriatua and Aquila ehryea7toe. The greatest range of interspecific (or intraspecific?) variation is found in the hook- Uiftea kites of the genus Chondnohi'erau (fig. Z}L).

'""Q>

Fig. 20t. Bill form in the genus Chondrohienar. A. C. uneinatust B. C. uilsoni.i , C. C, megarhYnehus. The upper part of the bill usually is smqgthly sheathed fron cere to tiir; but, in Gypaiitus, there is an indication of a suture fron the nostril toward the tip (fig. 202) , a l"ine similar to

Fig. 202. Bill and cere region of Gypa7tus barbatus. that observed in many other kinds of birds . This crevasse is not well marked in all specimens of Gypa?tue, but is distinguishable in most. Maehaerhamphue is unique in that the culnen is thin and drawn upwarcl. FALCONIFORMES 97(297 )

The tomiun varies from straight to rounded, "toothed" or notched. Apieeda and Hanpagus have two "teeth" while letinia }:'as a single FaLco-1ike "tooth". The tomiun is subject to heavy wear' like ihe tip of the bi1l, and as a result shows a wide range of variation in some species. Notched tomia nay be nothing but a broken edge while curved tomia may occur in species which usually have straight edges. Toothing of the toniun is correlated with the underlying bone structure in Ealgo, or the production of a thicker settion in the bill sheath which wears out to form the tttoothtt. The bill is usuaLly bluish black or horn colored. It is yellow in the larger species of Hali.aeetue and Torgoe traeheli'otus. The African subspecies- of Mi.Lt:usmignane have a yellow bill in contrast to the black one of the typical subspecies. AquiLa oerneausi and A. audas show some yeLlow and orange toward the base and tomium as do nany species of some genera (Butaetun, etc.). Color in the bill is rel"ated to food and shows seasonal as well as individual and geographic variationr 4s does the eye (JolLie, rs47). The accipitrid cere is usually yel1ow but ranges to bLuish gtay in Pandion. It is a soft, somewhat inflated region covered by flakey skin with a waxy finish. In GypaUtua (fig. 202), the cere is largely covered with bristLes, except imnediately around and anterior to the nostril; in this way it is like that of the falconid. In aLl other accipitrids, the cere is naked. The nos- triL is set in the cere and varies from a horizontaL slit, over- hung by a lip in Pandion to a round or vertically ovoid opening. The accipitrid cere fades into the skin of the bristled lores. In sone of the pernins and in Maehaerhamphue,the loral region is covered with numerous, srnal1, pennaceous feathers. Transitional stages between the plumed and bristled lores are shown by other pernins. The bill of the falconid varies from the quail-like one of Miloago to the short, strongly-hooked one of Mienaetur or Faleo. Falco is peculiar in its shortenedr pug face and resultant nodification of the vestibular folds. The rrtoothfr of the bill of Faleo is lacking in the Polyborinae and Herpetotherinae; it is almost natched by that of Ietinia. The bill may be whitish, yellowish, or horn as in the accipitrids. The cere is like that of the accipitrid in coloration but somewhat distinct in forn and usually covered by bristles. The bristles are least evident in Faleo and Polybonus. The nostril tends to be round or bluntLy triangular and rimned by an ele- vation of skin (except in the Herpetotherinae where only the dorsal nargin projects). Polybonue is the extreme in nostril forn (fig. 1"188). The vestibular fold shows at the nostril opening; its tip appears as a distinct central tubercle in Faleo. The lores are always bristled. Sushkin (1905:63) described the horny palate of the falcon as ridged whereas that of the accipitrid is not. A palatal ridge is present in all of the falconiforms but it is most strongly and unifornly developed in the falconids (figs. 60, 107). Sagittariua agrees with the accipitrid in most details: the cere is yellow, the nostril ovoid (fig. 79A), the lores are 98(298) M. JOLLTE covered with bristles. However, a vestibular fold is lacking. The outline of the feather nargin on the nandible was identified by Sushkin (1905:63) as characteristic. In the accipitrid, the side of the ramus is feathered to the horny sheath but the ventral edge is unfeathered for a short distance posteriorly; thus there-is a distinct indentation or line between the lateral feathers and those of the throat. Such an indentation is lacking in the Falconinae and Polihieracinae but is present in the Herpetotherinae. The reduced face and chin feathering of the Polyborinae- obscurs it. Many accipitrids have crests on the head, and similar crests appear in many-groups of birds. The form is quite variable and within a genus (Hieraa7tue) a crest may be present or absent. Any evidence of a crest is lacking in the cathartid (head feather- itb reduced also in the natal. stage) whereas Sagittanius has loig crest feathers which hang well down onto the neck. None of the-falconids has a well-marked crest, although Henpetothez'es has a broad nuchal one. Amongvulturine species, the head feathering is stronglY reduced. The face may havb brightly colored areas or be uniformly colored red or biackish in the vulturine species. The most highly rnodified head cover is that of Sareonamphus. Torgos has large lateral lappets of skin on the neck. Vultur has wattles at the chin and a- iarge fleshy caruncle on the top of the head in the adult. Sagittariue is- unique in its_ _we11-developegeye lash-es. Micraetur ind Cireus have facial ruffs much like the owl. These seem to be correlated with a large external eaf at least in ' Cireue and the owls. Wing and Tail Proportions have been commonly used for differentiation: the length oi the wing as conpared with the tail or the length. of th" exp6sed part of the primaiy feathers and the- position of the tips of- the pi.imaries in ielation to the tail. These proportions are highly irariable. Such genera as Elanus, GampsonAr, Chelietini,a' Elino'i,d,es, Maehaerhamphis' Ietinia' and Harpagus are shaped like a pigeon. More hawklike but with pointed-wings (i:". t1P pli-- are l,li.Lrsus HaLiaslun and f etinaUtus .' Rounded, rnarils longes t) , " blunt-tipped wings are found in some kites (Rostrhamus'. Chondro- hierar" LZptod.on\ and some f,alconids- (Het'petotheres-, Mierastun) as weli as'being the node in nost of the accipitrids and all of the cathartids ind Sagi.ttarius. The long-winged forn (and wlf of flight) of pandioz cannot be matched in any other falconiforn but it is suggested bY Gypa'etus. A wing formula is based on which prirnary, feathel is longest and the deicending order of lengths of the others. Within one genus (Aquila), tf,e fornula may vary considerably; within a species i.t nay-n6t be constant (faleo rustieolus, Buteo poLAsoma). The notching of the primaries has been discussed as a functional feature bt Saville- (1957). As a taxonomic feature it has been used to separate the mainland subspecies of Buteo p,oLysoma fron Buteo poe.iloehrous. Notching is observed in all falconiforms (weak-est in Chelietinia) but it is also found in other birds as well. The tail ranges from deeply forked in Chelieti,nia and FALCONTFORMES 99(299)

Elanoidea to square or rounded; fron short in Ierathopi.us to long and floppy (1ike that of a nagpie) in Erythrotni.orehis, Eutriorohie, Driotriorehie, and urotniorehie. The tails of species of a single genus nay be square, rounded, or wedged (Euplectes--Delacour and Blanc, 1953-1934). The basic proportions of wing and tail may vary within a species (fox sparrow, Lins- dale, 1928). As a result prinariLy of variation in tail length' such cLosely related genera as Pernie and Elanoidee or Chondro- hierar differ narkedLy in proportions. The same is the case in conparing Roetthamue with HeLieoleatee, but otherwise they are so simiLar that they can be considered congeneric. On the other hand, the close resemblance of Chelietinia artd ELanoidee or Neophnon- and Necroeyntes appears to be due to convergence. The alnost exclusive use of body proportions has been the weakest point in systenatic work and has resulted in nany unnatural associatiorts (we11 illustrated by Swann's work). Hind Linb Sagittari,ue is unique in having very long, cranelike legs. Anong the accipitrids, the Legs nay be fairly long and thin (aeelpiter, Ciieus, Gymnogenya,-(stebhanoa'dtue,Genanoepiza, Hgponorphnus, etc-.) dr shbrter and thick Harpia, Pitheeophaga). The shortest legs in proportion to body length occur in the kites (Elanoidost Chondrohierax, Elanue, etc.). Arnongthe cathartids, the fossil Neocathartes had long legs (intermediate toward Sagi.ttariue) as did sone nore recent forms; the livin-g species have relatively short legs. The falconids are like the average accipitrid. Conpton (1958) pointed out that the thigh of the osprey and cathaitid is more closely attached to the trunk than in- other faLconiforms. This peculiarity is approached in Sagittariue- and the aegypiins. The length of the tibiotarsus has been compared with that of other leg and foot segments but an accurate measurement can only be based on the bone. The length of the tibial plunes has sometines been used (subspecies of AquiLa ehnyea7tos). In Sagittariue, the cathartids, and Pandiory, thesg do not extend beyond the heel; in nost accipitrids and falconids- they do. The shank is feathered at Least to the heel in all falconiforms. The length of the tarsometatarsus has been used for proportions within-the linb as well as compared with that of the wing, tail, or exposed prinaries. UtilLzation of such proportions mal reveal significant points at the species level (Friedmann, 1950). The proportions of the toes and claws are variable. Aeci- basis of the length- of the -rniaatepiter and- Aetur were separated on the toe as conpared with the outer toe. Analysis of this proportion indicated that there was every gradation between the short- andlong-toed species. Correlation of toe proportions (also as related to tarsonetatarsal length) with feeding habits has been nade in the cases of several species of Faleo (see Stresemann, L927-34:465-466). The evidence indicates the possibility of short- and long-toes species being parallel adaptive variations in several 1ines. Peculiar to certain accipitrid species is the fact that the outer toe is shorter than the inner (Ridgway, 1874:83-'fctinaltus, r00( 300) M. JOLLIE

GymnagenUs,and Genanoepiza). The cathartids stand apart in having a short and weak hind toe, while the anterior toes are relatively long and slin, indicating 1ittle grasping power. Sagittarius has short thick toes and a foot which compares well with that of the accipitrid. The toes of the falconid are proportionally Longer and thinner than in the typical accipitrid, but exceeded by some species of Aeeipiter which show "paralle1", bird-catching nodifications. The pads rnarking the penultinate joint of the anterior toes are better developed in falconids (matched in sone species of Aeeipi.ter, least marked and alnost cathartid-like in Polybonus). There is no absolute gap between the accipitrid and faLconid in toe details but the feet of most, if not all, species indicate to which group they beLong; the vulturine species are nost nearly alike' The foin of the claws is one of the primary features of the falconiforns, although the cathartids nust be considered exceptions. They range from blunt and straight (cathartids,..PoLy- borus) or long ana straight but sharply pointed (fetinaUtus) to well-curved and powerful weapons as in most species of accipitrids. The grooving on the under surface as well as the form bf the edges ia used taxononically. Ridgway (1874) and Fr:i.ednann, (1950) noted the lack of a ventral groove in the Elaninae. The cathartids and Sagittariue have well-developed membranes between the second and third and the third and fourth digits whereas the accipitrid and falconids have only the latter one. (Table ZZ). A nembrane is lacking in Pandion and is nuch reduced ti * ,t * *'rt rt * * * * * tr * * * *-rt tr * * * * * tt tt ,t tr * * ?t * tr Table 22. Squanation of the tarsometatarsus among falconiforms.

Cathart id finely scaled inner and outer membranes Sagi ttari id coarsely scutellate inner and outer menbraneS Accipitrid Pandion finely scaled no mernbrane Elanin finely scaled slight outer membrane Aviceda coarsely scaled slight thick outer mernbrane other genera of from scaled to outer mernbrane perninae weakly scutellate Machaerhanphus finely to coarsely outer membrane scaled Ic tinia scaled to scutellate outer membrane Milvus coarsely scaled to outer membrane scutellate Rostrhamus scutellate outer membrane Hal i aeetus scaled or scutellate outer rnembrane Icthyophaga coarsely scutellate outer membrane GypaEtus s caled outer membrane Neophron s caled outer mernbrane Circa€ t in s caled outer membrane Aegypiin scaled outer membrane Aqui 1in feathered all around outer mernbrane FALCONIFORMES ro1( 301)

Accipitrin (Buteo, scutellate to coarsely outer membrane Aeeipi.ten, etc.) scutelLate--some feathered Circus scutelLate outer membrane Falconid Herpetotheres scaled outer menbrane Micrastur scaled or scutellate outer menbrane Polyborin weakly scutellate outer membrane Polihieracin weakly scuteLLate outer membrane Falconin weakly scutel.late outer menbrane ** ** t * ** * * * i * * * * * * * i * * * * rt rt * * * * * * * in the eLanins and Auiceda, Anrongthe faLconiforns there is a continuous range of variation from-a finely reticulate (scaLed) cover of the tarsonetatarsus to scutellbte or nearly booted ones (Table 22 and fig. 203). Scutes eppear to be forrned by the fusion of scales of the reticulate-cover. The size and number of scutes is related to the size and thickness of the linb, and the anount of flexion and protection required. The primitive, fairly large bird, probably had a reticulate pattern which has been modified in snaller speciesr or where-ad,dedprotection was needed (Large ratites and some Long-legged types), to a partiaLl.y scutellate (or ocreate one- -Bee Boetticher, t929). The cathartid belongs to the primitive array of larg9 species with a reticulate cover. Sagi'ttari.ue (fig - ?04) has stutellate cover fore and aft; posteiiorLy the scutellae extend down nearly to the hind toe. The accipitrids range fron pel- fectly reticulate to alnost perfectly scutellate with some fusion of the scutes posteriorLy to- form an irnperfect ocrea. Within one species the extremes nay occur, as in the large species of HLliaeetue and in the slecies of Miltsue ot Ietinia, Although the pattern of scales oi scutes is variabl"e, thg falconids generally- have only three scutes above the toes on the anterior aspect of an otherwise reticulate tarsometatarsus. The tarsometatarsus is feathered in sone of the accipitrids. It is feathered all around in fctinaUtue' Aqui,La (Unoalituo), 0roaUtus, Lopha'etue Hi.enaa|tua, CaseinaLtue, Spi,za'dtue, Spi zaetur, " Stephanoa|-tue, PoLema'|tue (i.e., the aquilins), or it has a series of scutes up its posterior aspect in Buteo Lagop_us,-Buteo negali.e, and some ipecimi:ns of Buteo henilaeiua. The feathered taisometatarsi of the aquilins and species of Buteo are parallel developrnents from the prinitive scaled condition. The coloration of- the tarsometatarsus and foot is of no value. It ranges from bright red in sone species (Saneogype, and some species of taleo) to slate g\ay or-qlackish. The feet and lower leg of most accipitrids and falconids are yellow like the cere. Sagittarius has ltesh-colored legs and feet whereas those of the cathartids range fron pinkish white to b1ack. Light-colored legs (as the cere and the skin around the eyes) sornetines vary with ager the diet, sex, and tine of Year. Plunage Coloration The patterns of plumage coloration of the various types of falconiforns can be characterized (coloration of the face, bil1, and foot have already been nentioned). The cathartids tend to 102( 302) M JOLLIE

Fig. 203. Anterolateral views of tarsometatarsi and feet of A. Elanue eaeruleus, B. Auieeda euberistata, C. Miluus mignans, D. Buteo jamaieensis. Posterior views to right in C and D. FALCONIFORMES 103( 303)

Fig. ZO4. Anterolateral views of distal end of tarsometatarsi and feEt of A. Sagi,ttari,ue eerpentarius, B. Cariama cristata, and C. Chunga butmeisteni. have solid black plunages and resenble their reLatives in the Fetecanifornes ani proEeLLariifornes in this respect. The strong oaor-of skins of these orders also suggests interrelationship. The plumage pattern of Sarcoramphue i; the extreme in the reduc- yoPng black. The condors iiott'of pigrnLntation. However,--the ?Te , have whiie-areas in the wing. The cathartids are such a snall g"orrp i6at the plunage posslbilities can only be_assumed to paral- pelecanifor4s' iel ttrose-l observed ii the procellariiforns Or frypoihesis on the evolution of coloration can be stated. fne piiniti.r" coloration was pr-obably barred brown with or without irriiescence. From this thre-e basic types of nodifications arose: modification of pattern itself Qtiz'., stripes for bars), general aii"tio" of pigmbntation to white locally- or over. nuch of the b;at;-;i itrci"it"d pigmentation to solid-b1ack (which rnay then.be aiiiriea to white in^pirt). Conbinations of these have occurred aiitrough one or ttre bttrei is predominant. Black or black and white pifi.t;; thus represent an exireme just as does the light-colored or white plumage. Sagnituri,ie has a sirnilar col-or-pattern il )'oun8 and adult which cin be described as regional dilution and intensification oi color. There is no real similarity to the cathartid but only r04( 304) M. JOLLIE the suggestion that the two groups nay have stemmed from a uni- forrnly colored ancestor. The accipitrids and falconids are sinilar in having a primarily barred plumage pattern. The more finely-barred pattern seems to be the nore primitive; the tendency for reduction in the number and increase in width of bars in the adult'indicates the direction of nodification. A barred pattern may be lost in most parts of the plunage through intensification of pigmentation, but frequently appears at the base of the prinaries. The convergence between Peeudogyps or Irigonoceps and the condors should be noted. The large vulturine species (not including Gypohieraa) as a group and the species of ttaliaeetus and AquiLa show rnodification toward solid coloration. This is not evidence of commonancestry but rather a convergence that night have adaptive va1ue. GApo- hierar and Neophnon are an opposite extreme with largely white plumages. The "primitive, species of the accipitrids, the kites, frequently show highly rnodified plunages: the elanins might be chosen as the nost extreme, followed by some pernins. The most prinitive patterns are found in Buteo, Aeeipiten, and closely related genera suggesting their more recent radiation from a relatively unspecialized ancestral type. As would be expected, adult pl.unages are more modified than the juveni.le; in some species of the the prinitive core the juvenile pattern has shifted toward the rnodified adult type. In what appear to be advanced pLunages,thbre has been a loss of barring or at least a submergence of the bars into the generally darkened plumage. Where the number of bars has been reduced in certain species, sone feathers nay be nore advanced than others. As an example, in Aeeipdter badiue the central tail feathers may show four broad bbrs and the outer ones seven. Within one species there are different 1eve1s of pattern differentiation (Buteo Lineatus and Cireue aenuginoeus). The accipitrids as a group show certain peculiarities of plumage markings such as the broadly barred underparts (emphasis or intensification) of certain kites, Cinea|tue, etc; the median black throat stripe (Table 23) found in alnost all subgroups ex- * * * * * * * * * * * * * * * * * * * * * * fc tt * * tt f( tt tr * t( Table 23. Occurrence of a rnedian throat stripe among accipitrids. Probably variants of other species have been overlooked in com- piling this list. Machaerhanphusalcinus young and adult Aviceda jerdoni young and adult Aviceda subcristata young and adult Aviceda madagascariensis variants Leptodon palliatus young Pernis apivorus variants Harpagus bidentatus young and adult Harpagus diodon young and adult Accipiter trivirgatus young and adult Accipiter virgatus young Accipiter griseiceps young FALCONIFORMES 105( 305)

Accipiter hapLochrous young Accipiter eichhorni young Accipiter cooperii rare variant Urotriorchis macrourus young Dryotriorchis spectabiL is young and aduLt Spilornis rufipectus variants SpizaEtus alboniger young and adult Spiza6tus cirrhatus young Spiza6tus nipalensis variants KaupifaLco monogranmicus young and adul.t Butastur teesa young and adult Butastur indicus young and adult Butastur Liventer young Butastur rufipennis young and adult Buteo nitida young Buteo magnirostris young Buteo platypterus young *************** * * * * * * * * * * * tr * * tr * tt tt cept the Elaninae and the vuLturine species; the choice of color foi the barring of the underparts, usually reds and grays; and the tendency to gray and white pLunages. The falconids do not show the compLete accipitrid pattern, aLthough Mi,erastun is indistinguishable fron, and shows the same kinds 6f variations dsr Aeeipiien. The faLc6nids tend to have finer barring of the body plumage than the accipitrid and this barring shows paralle1 nodifications' Ridgway (L874275) re- narked-that in Mieraetun "The nunber of tail-bands cannot be relied on as a specific character, since there is 1 great variation in this respect among different individuals of the sane species. Thus, bemi.tonquatus has in some specimens on!)r four binds, and in others as many as seveni rufieolli,e has four to six; Leuoauehenhas six to seven, and eoncentricue three to four. The number is greater in the young pLurnagethan in the adul t . 'r Anong the falconids the nost p,rlmiliYe pattern appears-t9. be that oT a.t,Luago. Each of the subfanilies-shows stages of dif-- ferentiation from this type with extremes reached in Herpeto- theres, Polybonus eheriuay, and Poli,hieran eemitorquatua. The falccininae have sexuaL dirnorphism of pattern in sone species, general1-y a differentiated adult plunage, and all have a noustache mark or narks Both the accipitrid and falconid have striking exanples of paralle1 pattern deveLopment. The extrenely close similarity between Aeeipiter pectoralie and Spizg'dtue ornatus or between Faleo albigulanib and Faleo deiroleueue, along with the basic uniforrnity- (or conservatisrn) of pattern throughout thg accipitrids or falconids, suggests that only a few genetic factors are involved and that these renain in fairly stable combinations. Sunmary and Conclusions Eicternal characters, as indicated here, are no more conclusive than internal features in deternining the subdivisions of the subdivisions of the falconiforms, but like the internal 106( 306) M. JOLLTE features there is a suggestion of independent origin for the several groups. The cathartid shares the webbing of the toes with Sagittarius but also with nany other birds (fig, 204). Sagittarius differs in the scutellation of the tarsus, the forn of the foot and bill and in its general body form. The cathartid can be conpared better with selected aspects of the procellariiforns and pelecaniforms.- Cornparison of Sagittarius and the ciconiiform (preferably a stork) finds little agreement--even the plumage pattern cannot be fiirted effectively. The stork has a reticulate tarsus, which night not be considered of great inportance in some casesr- but heie it seens to be significant. The only conclusion is that there is no identifyable commonancestor for the cathartid and Sagittariid, certainly none which could be identified as belonging- to either of these types. The accipitrid and falconid types do not intergrade in any way suggestive of significant'conmon ancestry although a certain amount of convergence is evident, Mi,enastun and Hglpetotheres have been confus6d with the accipitrid but can be differentiated on the basis of the bristled cere and the form of the nostril rim. Details of the feet are useful but not conclusive. In any case, they appear to be specialized end lypes rather than relict prinitives in most of their external features. pandi,on differs markedly from the average accipitrid and represents an extreme in nostril details, plunage pattern, , sqirarnation of foot and toes, and ger.reral texture and size of the feathers. If this genus is identified as a specialized accipitrid, it supplies a-working lirnit for variation within such a type. In view of Pandion and on the basis of external features, the falconid and accipitrid could have a commonancestor although the degree of similarity (and lack of intergradation) does not necessitate this. PHYLOGENYWITI{IN THE FALCONIFORMGROUPS

Introductory Renarks Before- considering the question of the interrelationships of the four falconiform types described and evaluated in the previous sections or the place ilf each of these groups in a phylogenetic- scheme for birds, it is necessary to speculate on evolution within the class and within each of the types. The direction or trends of change that have occurred in- each structural unit should be defined so itrat comparisons will take on added meaning. The possibility for error is great, but the results of the examination of several anatomical systems should reinforce each otheri i.e., the sequence of events as indicated by pterylosis will support that suggested by the skeletal system (reciprocal illurnination of Hennig) . As an evolutionary starting point let us start with a hypo- thesized ancestral bird. To be cal1ed a birdr 8r organism must have certain features. Among them the following seem important: FALCONIFORMES 107( 307) feathered*; warm bl.ooded; four-chanbered heart**; lung of tubu- lar structure and air sacs which extend throughout the body; true bill with horny sheath, teeth lacking*:ttt; lacrimal , supraorbital' postorbital, epipterygoid, and ectopterygoid missing_ from skull; mesethnoid well developed; lateral ethnoid present; large orbitosphenoid (laterosphenoid or pleuros_phenoid) pre-sent; orbit large aird confluent with the antorbital fenestra in front (Ioss of LacrinaL?) and the tyrnpanic recess behind; loss'of contact between the squarnosal and-quadratojugal; relative enlargem,ent of craniun (related to increased activity and agility); mandible without ieeth (see footnote botton of page); presence of a neckelian fenestra and absence of a coronoid; neck relatively long' trunk short, and tail abbreviated; thoracic cavity enclosed by ribs articulati,ng with a ventral, keeled sternum; pel.vis 1?rg9l attached to a laige number of synsacral. vertebrae, i1iun, ischium and pubis extending posteriorly behind the acetabulunl anterior appendage nodified as wing, digils _II,_ III, ?nd Jy present, ai-git V-vestigial; hind limb welL developed_for bipedal stance; folr-toed fooi designed for perching or walking; a fused tibiotarsus in the shank, fibula reduced distalJ.y; tarsometatarsal components fused, first netatarsal developed,only distally-and- articulating with tarsonetatarsus; and, 1astly, many details of nyology, paiticularly the great de-velopmentof peetoralie, Any airinal'which does not fit ttris definition cannot properly be called a bird. For exanple, Arehaeopteryn should be viewed as too simiLar to the reptiie and not nbdif-ied enough to be a bird. This definition is not adequate, as regards many of the details described in this account, and needs to be enlarged upon. The following additions seem to be approPriale.- ^- Pteryloiis: feathers well developed and differentiated on nost part; of the body (showing specialized forns correlated with iunctions); aftershaft well developed; pterylosis pigeon- like, the ventral tract undivided on the breast; only the protegnen present; 11 prinaries in the wingr-9llpaf.remex with tovert preient, wing di.astataxic (Steiner, -1918): , Skirtt: pitate-dromaeognathous (upper bill involving both prenaxilla and rnaxilla; the latter more so than in the ^Neognathae;maxillopalatines not touching,at the midline but in conlact with the vomer, which is flattened in the plane of the palatal surface; median nares to either side of the posterior con- ilortion of the vomer; pterygoid large, a unit structurer -in tact with the vomer anteriorly and supported nedially and near its posterior end by a large basipterygoid Proce-ss; palatines relatively srnall); prefrontal well developed with a distinct supraorbital processl palatines relatively,smalL) ; prefrontal weil developed with a distinct supraorbital process; nasal opening holorhinal or incisorhinal (neW term--the opening forms an acute angle dorsoposteriorlY). ***t*****-*****'*********.********** * pennaceous structure gives supporting area for flight. ** left systenic arch lost. *** teeth of Heeperonnie may be a case of neoteny, an adaptation for fish capture torrelated with loss of flight and-diving habit; rcthyonnia *"y also be a case of adaptation fof fishing. 1oB( 308) M. JOLLIE

Mandible: anterior splenial fused with dentary (see, Jol1ie, 1957a); no (?) coronoid. Body skeleton: L4-L5 cervicals; 9 vertebrae between nain brachial and pelvic nerve roots; long synsacrum of 10-15 vertebrae; short caudaL series of about seven units; ribs well developed with uncinate processes; sternum large with deep keel, two deep incisions in posterior nargin; coracoid with simple head and distinct distal expansion at articulation with sternum, snall procoracoid process; clavicles fused ventraLLy, sma11 round head in contact with coracoid and a thin extension extending back to contact scapula; scapula thin and blade-1ike, pointed posteriorly; pelvis formed of iliurn, ischiun and pubis fused at the acetabulurn, separate behind this vertical; iliurn extending length of pelvis, attached to the transverse processes of the synsacral vertebrae; no prepubis.- Limbs: claws on second (most anterior) and third digits of wing; two bones in the carpal region with a sesamoid anteriorly for the tensor patagii longus tendon. Myology: belly of eeratoglossus lies above insertion of transuersus hyoideus; dernal musCles weakly developed; tt'acheo' bronehialis inserts on second and/or third bronchial semi-ring; eopansor seeunday,iorum present; Latissimus dorsi with two parts; tenes major small; tenes minor smal1, arising from the anterior margin of the scapula (and clavicle?); deltoid two parted, one part serving the propatagium and its tensor patagii breuis tendon diffuse (not distinct bands); bieeps slip to propatagium present tt,ieeps with pa"s anillaris; peetoraLis a single nass wittr outer and deep insertions; tensor facia Latae arising along entire dorsal iliac nargin; reetus femoris (ambiens) present and inserting through patellar sheath into superficial digital flexor mass; bieeps femoris et semimembra.nosus inserts through "aeeeesorius"; eaudiLiofemonali,s with caudal and iliac origins; plantarie two parted, the pars tibiaLis incorporated into the gastroenemius s.heet; soLeus present and well developed; f L,eaor atgttorum supez.fieiaLis subdivided into five perforans and perf oratus p-arts; the deep f lexor tendons fused throughout much bf tne length of the tarsonetatarsus; tibiaLis antev,ior covered largely by the pe?onaeus Longus. viseral Anatomy: gut 1-ooped and fairly long; stomach simple (as in the falconiforms); caeca present but sma1l; only the teft ovary and oviduct developed (perhaps vestige of right)l testes 9qua1ly developed; both carotid arteries developed-and together in th9 hypapophysial canal (Glenny, l-9ss), vestiges of ot[er vessels in the region of the heart; syrinyx of simple tracheobronchial type (much as in the accipitrid) with slight medial and. lateral tynpaniform membranes; ful1 nurnber of pulrnonary air sacs present anq peletrating bones; air spaces of head (froln nasal passage and middle ear) also present. External anatomy: in proportion and size about like Areheop-te_t'yr; tail shortened and composed of Lz feathers; form adapted fof easy progression (walking) on the ground, or junping from branch to branch in the-tree; wing not utitized-in ciimfin! other than as a balancer; bill somewhai longer than deep and hooked at the tip; covered by a plated rhamlhotheca composed of FALCONIFORMES 109( 309) four or more pieces above and seven below (Jol1ie, 1961); nostril nidway along the length of the bill; sirnpLe vestibular fold and slightly rolled nasal tubinal; olfactory chamber with snall fold or rolLed tubinal; plunage well developed and with smooth outer surface, brown coLored, with darker cross bands; foot and tarsometatarsus covered with fine scales (reticulate). The prinitive birdr BS described here, did not differ narked- Ly fron Living types except for its internediate condition in some anatonical features. If this is an accurate description of the avian archetype we can conclude that most f.iving groups (fanilies in some cases, orders in others) have all progressed equally far fron this starting point in one feature or another. Not bne of the existing groups can be considered as more primitive than others unless comparisons are based upon rpecific features. The ostrich is nore primitive than nany types in terms of its large basipterygoid processes but nore modified in the large gap between pterygoid and vomer In order to consider the direction of evolution within the falconiform groups, we must have an understanding of the basic adaptive radiation within birds. For this we can hypothesize that the prinitive body form was retained in some birds but underwent rnodification in others along three basic lines: (1) cursorial habit coupled frequently with increase in size and loss of power of flighti (2) swinning habit involving pterylotic changes, webbed feet, and slightly increased size; (3) arboreal habit involving shortening of leg and no change or a decrease in size. These basic routes of hind linb specialization were accom- panied by nodifications in flight. An aquatic type couLd become flightl"ess (Hespenopnia or Speniscidae), or highly modified for flight (frigate-bird). These loconotor speciaLizations were prilnarily for securing food and are also reflected in bill, head, and neck forn. Once started on one of these paths there was some changing of directioni i.e., a swimmer could become a wader or take to the shore to become distinctly cursorial. Birds of an aquatic habit (swimmers, pelagic sea birds,- waders, and shore or marsh species) were most abundant in the period of the first radiation of birds because of the greater- ibundance of food. The land environment did offer possibilities for scavengers and predators. At first food included snall rep- til,es, the eggs or young of larger forns, mammalsor leftovers from a dinosaurfs meal while along the shores carrion was always' available; later with the appearance of flowering plants and insects a much more varied food supply (including fruits) was available and radiation of small species (particularly passerii forms) to expoit fully these sources followed. The Cathartid (Cathartoid) Type The primitive cathartid modified the ancestral pterylosis by adding a metategmen and reducing or losing the aftershaft. A solid plunage coloration of a dark brownish black was achieved before the origin of this group and was continued with little change throughout its history. The palate was schizognathous (in the true sense of the term), the naxillopalatines were well 110( 3r0) M. JOLLIE developed and spoon-1ike, the medial surface being nearly vertical. The prefrontal was free, basipterygoid processes were present, and the pterygoid developed a joint at its midpoint as a result of fusion with the palatine. The cranium was less rounded than in living types, the upper bill somewhat shorter. The vestibule opened widely through the external naris and the vestibular fold was reduced or lost. The body skeleton was fairly primitive, but the coracoid was strengthened as a result of increase in size; the sternum had two posterior emarginations. In their nyology this group was nodified. The peetoraLis became divided into two separate layers. (A tendency for this complete division is shared with the entire swimning line suggested above.) Also the teres minor developed a pq.?s internus. Neoeathartes gnallator approaches the archetype of this groupr 8r archetype quite similar to that of the stercorariid' procellariiforn, pelecaniforn, ciconiiform, and sagittariid types. The main doubt as to the propriety of this association stems fron the fact that the first three groups have a webbed foot whereas the others do not. The reduced hind toe of the cathartid suggests that the foot may have been webbed, but this specialization was lost with the move to the shore and land where they continued their scavenging way of life. Perhaps Neoeathantes had longer.legs and shorter wings than the prototype, since these could be specializations for shore- living. It is possible that the swimrning prototype may have passed through a vulturine wading stage toward its present habit. ttre possibility of a fossil "cormorantrf being present aLready in the Cretaceous would argue against a basal position for this Eocene fossil (ltretmore L951b:53) . In any case , Neocathartes is probably nuch like the ancestor of the present cathartid species. 0f the living and well-known fossil species, Cathantes and Coragyps retain more of the primitive features. In these, it can be assumed that the bill is elongated, head feathering reduced, external naris lost (along with the vestibular fold) prefrontal fused to the orbit margin, the cranium more rounded,' and nost units of the body skeleton retaining a suggestion of the ancestral, swimming proportions. The condors showed marked specialization in terms of size, in pterylosis, in plurnage and body coloration. In this group the greatly elongated rostrum, the proninent exoccipital and basiparasphenoid processes, the form of the cranium and its superficial overlay of bone along with its relationship to the neck all denote specialization. Saz,eoramphus is intermediate between Cathartes and Vultur in nany respects (but not in plunage coloration) . Gymnog7ps is more modified than Vultur, and its trend in skul1 modification is carried even further in BneagAps. Morphologically the fossil Gynnogyps amplus (Fisher' L947) and tiving Gymnogyps ealifornianus intergrade and probably represent subspecies in time. (Howard, L947, expressed the same opinion. ) The variability of the fossil forn is explained by the long period during which the remains were deposited. It is probable that many of the tar pits were open and actively collecting at widely separate tines and, in the opinion of Howard (conversa- tion), for only relatively short times. Fisher (1947) stressed FALCONIFORMES rr1 ( 311) the gap in measurements, but the gap exists only when averages are considered; further, the nodern population is small and night conceivably show a reduced range of variability because of strong environmental seLection on a dying species. The evolution of the cathartidae (fig. 205) , including the phasmagyps Pal aeogyp s (Neocathart i dae) Neocathartes Cathartes

Coragyps q)

+J Sarcoranphus tr

+J Vultur 'r{ f1 Gymnogyps Breagyps

Teratorni s (Teratorni thidae )

Fig. 205. The phylogeny of the cathartids .

Oligocene Phasnogyps patritus and Palaeog.Aps_.pro_dromus, has been schenatically illustrated by Fisher (1944:294, fig. 47). He indicated (p. 293) that Bneagyps elanki was ancestral to GAn- nogApe and VuLtun. Such a view is hardly tenable inasnuch as Brbagype was a contemporary of Gymnogyps and as such could not be anl-estral (see Howlrd, 1952). Bneagyps with its much elongated rostrun, is a more extreme specialization than Gymnoggps_and is thus even further from the ancestral type. The amount of change evident in conparing Pleistocene fossils with modern forms suggests that divergence of the condors took place in the Plio- cene. The most extreme cathartid in some respects is Tenatornis merriami,. The rostrum is highly modified, the interorbital septum is inperforate, and the olfactory capsule reduced in size. The pectoral girdle and the sternum are modified in parallel fashion to those of Ftegata and Dionedia. In its pterylosis_altd myol ogy Ieratonnis nay have gone further than the condors. This giant-vulture also retains certain unmodified features, such as ttre configuration of the cranium and quadrate and the tarsomrneta- tarsus. The retention of such feattires suggests an early divergence of this line from that giving rise to the living cathart ids . Teratonnis ineredibilis (Howard, 1952) was described from a "cuneiform" (radiale) bone. It does not seem worthwhile to specu- 112( 312) M. JOLLIE

late on the nature of a species known fron a single wing element. Thus, comparison of the condors, or Tenatornis , with other types of birds can only be rnisleading since, as the most specialized cathartids, they can resenble other groups only through convergence. European fossils which have been identified as belonging with this group are too poorly known to be allocated. Cra.craft and Rich (L972) commented on the Tertiary (Eocene, 0ligocene of France) occurrence of European cathartids, Eoeathantes, Plesiocathartes" Diatropornis (Tapinopue), and Amphiserpentarius. Relative to the first of these they conclude (p. 273) that several features "strongly suggest that Eoeathartes is a cathartid vulture. Un- fortunateLy, neither Lambrecht's discussion nor his figures is sufficientl"y detailed to construct a generic diagnosis." Wetmore (L955:54) reached a sinilar conclusion. PLesioeathartes europaeus (Gaillard, 19C8:4L-44) is based on a part of the shaft and the distal trochleae of the tarsometatarsus. In size this fossil is much smaller than known cathartids, and in details of structure shows nothing that can be used to link it conclusively with this group. It is of a generalized form suggestive of several groups of birds. The hypotarsus, which might irnprove the identification, is nissing. Diatropornis eLLi,ott, from the same formation as PLesi,o- eathartes (Gail1ard, 1908:48-51, fig. 8), is known fron the complete tarsometatarus and the distal end of the tibiotarsus. It probably is a cathartid, but identification as such needs con- firrnation from other parts of the skeleton. Cracraft and Rich (L977,:275) identified the tibiotarsal fragnent of Amphise"pentarius sehlosseri (Gai1lard' 1908:45-47, fig. 7) as a cathartid but again the evidence is inconclusive. Failure to identify these fossils positively as cathartids limits the range of this group to the New World and suggests that it developed in a fashion paralleled by an Eurasian counterpart, the aegypiins. Should these fossil genera prove to be cathartids (Diatropornis probably is) the probability renains high that they entered Europe from North America since here is where most of cathartid evolution appears to have occurred.

The Sagittariid Type This monotypic genus has no real fossil history. In terms of this study, AmynoptiLon robustum (Milne-Edwards, 186707L) does not compare well enough with the living representative to be considered related (see fig. L27) . It can be assumed that the ancestor of Sagittanius had the foilowing features: maxillopalatine processes fused at nidline, vomer a vertical plate between the internal nares, basipterygoid processes well developed and functional; prefrontal with large supraorbital process but no supraciliary, the body skeleton and limbs nuch like those of the ancestral cathartid except the foot modified for clutching, myology also like the ancestral cathartid, pterylosis unmodified from the primitive avian, metategmen and aftershaft present. The modern species cornpared with the ancestor, is nodified in the following manner: crane-like form with its attendant nodi- FALCONIFORMES 113(313)

fications of osteology (forn of sternum or pelvis--shared with Cari.amaand storks); pterylosis (i.e., separation of axilLary and sternal tracts, a trend indicated in the storks); specialized color pattern; feather crest; and scutell-ate tarsometatarsus. This genus probably represents an independent line establish- ed at the first radiation of birds. A wading (or cursorial type) with legs of medium length gave rise to two lines, one of which acquired a Long stabbing bill (the ciconiiform) the other which developed a shorter, hooked bi1.1 (sagittariid) and a foot modified for capture of srnall vertebrates by grasping. The Latter line gradually developed a more-and-more raptorial appearance. Along withr or after, the acquisition of the bill type, the legs became progressively more stilt-like as an adaptation for a particul.ar style of feeding. This history thus Presumes a relationship with the Ciconiiformes' one which involves some degree of paralleL devel"opmentwith the Ciconiidae. The Accipitrid Type This type is- represented in the Upper -Eocene by the tarsonetatarsus of.'balaeohibtae gervai.ei (Milne-Edwards , 1869-1871) , but it is not until the Oligocene renains of Paleoplaneus sternbengi (Wetmore, 1933) ot Buteo gnangeni, (Wetnore and Case, 1934) that other parts of the skeleton, including the skul1, are_repre- sented. The fossil renains fit welL into the range of variation shown by Living species; there is no evidence of transition with any other type of bird. Fron the fossils and from the comparative anatomy of the acciptrid group we can describe the archetype as follows: hooked upper bill, perhaps longer than that of the average living_ species; the-craniun rounded and fairLy large; supraorbital piocess of prefrontal large (and ? with a superciliary); the palate weakly desnognathous; the vomer a vertical plate between ttre internal nares; the basipterygoid processes vestigial; the pterygoid jointed at its niddle, the anterio,r part fused to the lalatine; ttre mandible without a meckelian fenestra; the coracoid strengthened and expanded diStally; the head of the clavicle drawn-back; the steinun with a single pair of incisions; the pelvis strengthened through further fusion and covering over of ttre spaces between the transverse processes of the synsacral vertebrae; the foot and tarsometatarsus altered for clenching food, the latter with two calcaneal processes separated by a deep flexor groove (1ike Gypohieras oT PaLaeohi,erau); ventral feather tract altered by the slight separation of an axillary band; a netategmen present and the aftershaft retained; prinitive plumage color pattern of bars altered dorsally and in the wings and tAiI by reduction in the number of cross bars and strengthening of their pignentation; the under surface lighter with some feathers cross-barred or striped. The history of the group can be described as one involving a direct conversion of the ancestral type toward a more specialized predator (in opposition to the view of a close relationship witit the Ciconiiformes). The ancestral type retained its walking habit, but increased its power of flight. Its adaptations to the original food source, carrion, involved developnent of a shorter, 111{(314) M. JOTLIE stronger, hooked bilL and grasping feet (both serving in tearing up food for swallowing). This typg of bill. and foot was a pieadaptation for increasing vulturine efficiency and aLso for seizing and ki1.1"ing snaller animals, including birds. Active redation was quickly foLLowed by radiation away from the scavenger ancestraL type. The Latter was tepresented at.any time in the history of the group by only a smal.l number of species, since this way of tife couLd suppl.y onLy a smaLL nunber of niche speciaLLza- tions, and there was little geographicaL isoLation within Eurasia- Africa for such birds. The radiation of this group from the anceslral type can be approached only through the existing species. Various genera can Ub'finked with'each oiher af" several livels of differentiation, some of which nieht be identified as subfarnilies, others as species cores or supergeiera. The trunk and main branches of the phylogenetic tree can only be surmiged. The best qppro-ach seems to 6tart from the wel"l-defined generic groups and work back through time, showing the different Level"s of interrelationship as fqr as is possiblo.- The Aegypiinae offer a starting point since they repr-esent the assumed-brimitive way of life (fig. 206). ActualLy' the very large species of this array must be considered as _highly specialized in- terrns of size and pLumage coLoration. Unfortunately' these vul"turine species are not anatomicaLLy weLl known in spite of the fact that aff of then have been dissected at one tine or another. It is evident from the study of osteoLogy, however, that the present subfanil"y Aegypiinae (Pete-rs, 1931) it an undefinabLe array within which there is a weLl-narked core of species, the aegypiins (aee p, 50). The aegypiins probabLy radiated during the Pliocene into the present group of species. This radiation occurred as a result of ttre abuniiancb of large herbivores. The group is characterized by having an ossified lateral vestibular waLl. which outlines the external naris; in lacking a superciliary; in having a weLl.-de- velopedr pointed supraorbital process; and in sharing detail.s of coracoid, sternum, tarsonetatarsus, and pterylosis. The nembers of this group physically resenble, in sone respects, the Large eagles of the genus Aqui.La but this could be the result of need for strength in tearing up the bodies of the large nannals on which they feed. Atributing this simiLarity to convergence is in agreement with the conclusion that Aqui.La is a genus of some age and modification. The aegypiins can be placed in three supergenera. The first includes two pairs of generaz Torgos (traoheli,otue) and Aegypi.ue (monaehus), triganaeepa (oecipi.tali,e) and Sancogype (eatuue) . The first pair of species differs markedly fron one another in coloration and in detail.s of externaL appearance, such as the shape of the nostrils or the Lappets'of the former; in its skulL Torgoe has o more roundett cranium, the supraorbitaL process of the prefrontal bent tlown more into the orbit, the antotbitaL process heavier, the bill higher and more inflated, the nasal process of the premaxiLla wider and the maxiLLopaLatine processes sonewhat more fused at the midline. In spite of the narked differences, FALCONIFORMES 115( 315)

.PiIithnecojph -ac naaga

Haaarp?v{o)p:siisS

EuuJtl'ET:riLo)r(chriis

DrrioLolrt!rri

'ciirc:cz:arti-itr us; Halem)m€nart:orrnriLs I c ircaiit in Te>Ta:a1rt:hlotrpii.urs )

To)I$t:gcIo)S Ae)gyltyEtpri,usISS

'T'r'i .gcg,lor)nloc)c€ce.)p:)s 0.) ar{ Salrc('cc:olrg,iytr.,psps aegypi in .Fl .Fl Ne,)cr(:ro'ol,s.iyr'rtrtr.e:fS X Psrieu(udtdr.olrgyiYprp:)S O Gyt'ps

Gyt'paiaE.eltrUSt5)

'NeNe

GyGylporohhiirerra:a)x

mi.1v:vi'ir.n:iS;

I LAta J1iaiaa€e(ettu:us5 1haliaeetin IclhththhyYcODphrhaa8gaa )

Gymnogenys

Fig. 206. The phylogeny of the aegypiins and somewhat similar genera. the degree of sinilarity is such that one must conclude that these genera are geographic representatives of a not-too-ancient species, which in its appearance was closer to Aegypius. The same sort of relationship exists between the species of Sancoggps and Trigonoeeps. The former is more primitive in coloration but more specialized in osteological detail. The differences between these species are of a greater magnitude 116( 316) M. JOLLIE than those between the larger vultures. These two pairs of species appear to be derivatives of a common ancestor which existed in the not-too-distant past. The tine of separation night be indicated by making these species congeneric; certainly placing them in a single supergenus is not out of order. The second supergenus contains the species of Pseudogyps and GAps. GAps is the more modified type. The elongated bill, schizognathous form of palate, reduction of voner, shape of the cranium, elongated basiparasphenoid and exoccipital processes, orientation of skull on neck (angle of base and back of skull), elongated neck and shortened trunk all denote specialization (see fig. 97 for convergence to Ggmnogyps). The dilution of coloration can also be cited. The third supergenus contains only Neenosyrtes monachus. The exact affinities of Neenoiyrtes cannot be stated, but it appears to be a member in good standing of the small-headed vulture group, which also contains the supergenus GAps. In coloration there has been intensification of general tone with submergence of barring; the head form is highly altered but the size and general- proportions are primitive. Neophron ana Neerosyrtes possess certain superficial resem- blances but disagree in detail. The former lacks the vestibular ossifications, has a reduced supraorbital process of the prefrontal, Btr ossified voner and a perforate interorbital septum. The base of the skull is somewhat inflated and differs in details. The sterna are unlike, and there is a lack of fundamental agreement in details of the tarsometatarsus. Details of the ventral feather tract separate these genera as does the plumage coloration of the adult (highly rnodif ied from the blackish, prirnitive plumage of the imnature). Neophron is "kite-like" in many respeLtsi while Neerosgt,tes compares better with the other aegypiins . Assurning sbrne kind of close relationship, such dis - agibbnent is unexpected; assuming convergence, it is to be The latter view has been adopted here. expected.- To make the problem of the relationship between Neoph,non and Neerosyntes even more interesting, a third fossil genus (fron North funerica) must be considered. Neophrontops amev'ieanus is more like Neophron than the aegypiin. These three genera are similar in general form with the cathartid, Coragyps sugges!ing ' that the shape of the head and bill has a strong adaptive value for a specific vulturine niche Tha other species currently included in the subfamily by Peters (1931) have no apparent close relationship. GypaUtus , barbatus is a widespread Eurasian-African relict, probably of considerable age and highly nodified in terms of size, ptilosis, and plumage coloration. It differs so narkedly from_the_aegypiin type that only convergence can be assumed for those few features they share. Gypohie?ar angolensis could be a very old specigt r descended with fittfe change from the Cretaceous or Eocene accipitrid (based on similarity to PaLaeohtenat or Palaeoborus umbnosus, 8ee tioward, 1932:70-7I). The white coloration cf the adult (nodified frorn the blackish juvenal plumage) suggests great age or a long FALCONIFORMES 117( 317)

independent history. In its plunage coloration and stages, Gypohi,eras agrees with Neophron pe"enopterus . Since Gypa7tue" Gypohi.eran" and Neophron cannot be positively associated with each other or with any other species groups, they are considered here as independent lines going back to the first radiation of the accipitrid type. ALso of uncertain relationship are the haliaeetins (Ietin- a'dtue, Haliaeetue, and fehthyophaga). These eagles are probably closely related to the vulturins, but such a tie is as much speculation as anything else. The circa6tins resenbtre the aegypiin in having the lateral vestibular wal1 ossified, the calcaneal processes on a medial ridge, a rectilinear pelvis, and a reticulate tarsus. They disagree primarily in having a superciliary. Terathopius is in many respects internediate between the circadtins and the aegypiins, suggesting a close tie among these generic cores. The problem is one of whether all should be in one subfamily, or two. If two, the disposition of Tenathopius is difficult to decide. The circadtin core fray, or nay not, include Driotriorchis speetabilis and Eutriot,chie astur. The former has a fairly advanced style of coloration whereas the latter is quite primitive in this respect. Both have a finely scaled tarsus and a modified shape (long tail), particularLy the latter. Both appear to be very old species, or at least they are very different frorn any other kind of hawk. In this case, and in many of those yet to be discussed, there are no physical features known at the present, other than general appearance, which can be used to support relationship. Another core of species, based upon one feature and little else, is that of the aquilins (fig. 207). The feathered tarsonetatarsus (feathered all around) marks this group, the species of which at present are distributed among a dozen genera. Many of the species are large and powerful and in terms of proportions are distinct from one another. 0steologically the group is marked by the strong contours of the skull, the strong bill, and (usually) the flat rather than folded bridge of bone over the olfactory nerve. The body skeleton is powerful and relatively uniforn within the group; the tarsometatarsus is particular1-y characteristic (fig. 131). The pterylosis of this group is narked by the presence of small pennaceous feathers on the crop region. In coloration there is a trend among the smaller species for' a white undersurface marked with streaks or spots. The smaller species appear to be the less modified. Figure 207 shows the genera included in this array and their hypothetical interrelationships in time. The genus Spiza7tus is shown as having two parts, in the New and 01d World, which resemble each other as much through convergence, as through common ancestry. Very close to the aquilin group is the accipitrin, which is made up of three cores or supergenera: buteogallin, buteonin, and accipitrin (fig. 208). These three cores, along with several genera, which cannot be more exactly placed, share a more recent ancestor than tl'rat which gave rise to the aquilin. 0f this array the Central and South American species of the buteogallin super- 118( 318) M. JOLLIE

Italiaeetin

other vulturines

aegyp i in circatitin

Aqui 1 a

Aquila wahlbergi o cd H IctinaEtus .r.{ F{ LophaEtus (O1d World) Spiza6tus q Po1 emaEtus F{ 'Fl Stephanoa6tus (! 0roadtus Spiza6tus (New World) Spizadtus grinnelli Neogyps Cass inaEtus Hie raadtus Accipitrin

Fig. 207. The phylogeny of the aquilins in relation to sonewhat sirnilar genera and generic coTes.

genus appear to retain more of the pri_nitive features, a state- ment that is more of a hope than a defensible hypothesis. Probably the usual trend in this array has been a reduction in size with the smallest species (of Aceipiter and Auteo) being the more reccntly develoPed The buteogallin core is well narked in terms of plunage pattern, the long tarsornetatarsus, and other osteological details. This group includes Heterospizias meridionalis, a species usually placed closer to Aeeipiter than Buteo. The tarsometatarsus of this group is like that of ritanohierae gLouenaLleni and Calohierai quadr.atus fron the recent cave deposits of the Bahama Islands (Wetmore, 1957) or Hypomorphnus eneeta from the Middle I'liocene of Nebraska (Wetmore , L923) . The buteonin core intergrades with the buteogallin and the FALCONIFORMES 119( 319) Harpia aqui 1 ins l I Morphnus I I Harpyhaliaetus | . c) Urubitornis ( : /-t .F Hypomorphnus I S. t{ lo +) Heterospizias g .Fl \ .Fl t5 U Buteogallus I o I Busarellus J ri .r-l Buteo \ Kaupif alco I .5 tf.{ Butastur I tr t3 U) 'Leucopternis \E l-o . r'l H Parabuteo -/ P

'Geranospiza -',, .Fl I U tP U Accipiter |'L Urbtriorchis - [)s 'T 'Erythrotriorchis d I Circus -."

Milvus \ lc 'l -Lophictinia t .Hamirostra ( E $. o 'Hal ias tur -" 'Harpagus ., .r{ F : 'I ct inia t I 'elanins I 'Helicolestes ? I 'Rostrhamus i I \l'{achaerhamphus Pandion pern].ns ,,'

Fig. 208. The phylogeny of the accipitrin s in relation to thheir (?) next of kin. The elanins include Elanus' Gampsonyr and CheLietinia. r2o( 320) M. JOLLIE

accipitrin. It consists of shorter-1egged, broader- and longer- winged species. It is best developed in the New World, suggesting its origin here, but with a linited dispersal into the Old World where competition with members of the aquilin and accipitrin groups may have restrained it. The species of this core are intermediate in size but probably average srnal-ler than the ancestral form. Buteo fuseescens, the largest species, is probdbly larger than the ancestral type. To differentiate the buteonin core from that of Leeipiter, Erythnotriorehis, and tJrotv,iorchis is diff icult. 0n the basis of average proportions one can say that the accipitrin core species have shorter, narrower wings and a longer tail than those of the buteonin core, and the tarsometatarsus and toes are longer and slimmer. An average difference in shape of bill might also be cit'ed. In all features, hqwever, the two groups appear to intergrade perhaps due to adaptive convergence. 0f the accipitrin core species, Enythnotriorchis radiatus appears to preserve more of the primitive features of the group. In its proportions and coloration it is very close to Aeei- pitez, bingensi. ErAthrotriorehis dori.ae is less like Aeeipiten and can be compared with the milvins. The genus Cireus appears to be an early derivative of Aeci- piter in which the outer ear and the underlying bones have been greatly altered. The convergence to Mienastur in the facial ruff should be noted. The age of Cireus is well shown by the tendency toward pale gray plumages and sexual dinorphism of pattern and coloration in some species. The milvin genera (fig. 208) continue this series toward the kites. They share the reddish, cross-barred, or streaked plunage pattern with BusaneLLus" ButeogaLLus, Eutri.onehis, Erythrotriorehis, and some species of Circus. As the milvins cannot be properly defined, the inclusion of the species of Haliastnr. must be considered provisional. This group narks the transition from the scaled (primitive) to the scutellate tarsus found in nost of the members of this array. Somewhat more distant and showing strong speciaLization in size (small) and feeding habit are the other "kites". These might be placed in a kite subfamily, but this would be an artificial catch-al1. Differing least fron the accipitrin type are fetinia and Harpagus. These distinctly colored species with a notched tomium are close to the elanins in shape, size and plumage- coloration. The elanins mark an extreme of nodification from the prinitive type both in terms of plumage pattern (young and adult), and in rnyological details of the hind 1imb. Osteologically they are als.o distinctive in the separation of maxillopalatines, reduction of ossification in the nasal vestibules, and the forn of the tarsometatarsus. Rostrhamus and HeLieolestes are geographic representatives of a single ancestral type, highly modified in shape of the bill' body proportions, and in plumage coloration. Without detailed study it is only possible to guess as to the next of kin of these species but their placenent with other kites is probably not FALCONIFORMES 121( 32L)

incorrect. Maehaerhamphuais an extreme type without apparent close relatives. To ptace it closer to the elanin than to other kites is to assume that the open palate is of significance. Other sku1l features night be used to separate these. The distinctive features suggest a long and independent historl for this species. The peinin genera-The form the last and one of the most distinct species gioups. prefrontal is reduced, the supraorbital piocess is shortened,-and the superciliarl l-ost. The orbit is very l"arge with reduced nargins and all of the processes of the skuit are reduced. The tarsometatarsus varies but nay involve caLcaneal canals for the flexor tendons. Certain features of the sternum and pectoral girdle resenble thg aegypiin, i:€., _ prinitive. Some-of the detaits_ of nyology (the superficial flexor rnuscle of the shank and lack qf the reetus femoris) represent sirnilarities in nodification away'and fron the primitive. External bill form, foot tarsonetatarsal cover, plunage patterns, and the tendency for snal1 pennaceous feathers govering the lores and base of-bill mark this group. All of these features suggest- a long independent evolutionary-hist-ory, which night-be diagraned as in-figurb 209, Much of the resemblance between this group

Pandion Odontriorchis

Chondrohierax a Perni s tr tr t{ Aviceda o A Henicopernis I t , Elanoides ,'

other kites

Fig. 209. The phylogeny of the pernins and their next of kin. and the elanin (nodified plunage pattern' superficial flexor muscle details, and neetub femonie inserti.on) nay be due to parallel changes in the two lines. The-1ast, and most distinctive species of the accipitrid type, is Pandion hali.aetut. It usually has been placed in a sbiraiate family or a distinct suborder (see Table 2). This splcies dif f ers in its ptilos is , osteology,- lnd nyology so. diastically that it rnust either be excluded from the Accipitridae or recogni2ed as representing the extreme of allowable variation within i family. The similarity of Pandion to the pernin kites r22(322) M. JOLLIE

has frequently been remarked, but upon testing it is found to be very unconvincing. It seems best to leave Pandion as the sole representative of a. subfamily of great age but limited as to its biological success in terns of speciation. (The single species is cosmopolitan.except for South America at the present tine.) From the above discussion, most of the genera of the Acci- pitridae can be seen to fall into four large subfanilies: the Aegypiinae, Aquilinae, Accipitrinae, and Perninae. To this can be added the smal1 subfamilies Haliaeetinae, Elaninae, and Pan- dioninae. The Elaninae can be enlarged doubtfully, to include fetinia and Hanpagus and, even more doubtfully, Maehaerhamphus. ELanoides night be cited as interrnediate between eLanin and pernin. This leaves several genera standing alone t GymnogenAs, Gypohierax, Neophron, Gypadtus, Pltheeophaga" Harpyopeis, Eutriorehis, Driotri.onehis, and Maehaer'hamphus. Recognition of each of these as distinct sribfamilies is perhaps not proper, but for the present this is necessary. Sone of these genera (with the exceptions of Gymnogenys,Pitheeophaga, Hanpyopsis and Machaerhamphus)night be included with the Aegypiinae but to do so would soon draw in the kites and before long all of the accipitrids. Such a large nunber of subfamilies is not out of accord with the history of birds as described by Wetnore (1951-b:63). He suggests that birds were most abundant in the Miocene and Pliocene. During the ice age severe reductions occurred both in terms of numbers and of species. During the interglacial periods and after the ice retreated for the last tine there was some reradiation and an increase in number of species but also further extinctions. A dozen or so subfamilies does not seen too excessive when we consider that only remnants of a once widely diversified family are represented by living species. The Falconid Type The archetype of this group must be constructed without fossil assistance since it is a conclusion of this study that the Miocene Faleo namenta, described by Wetmore (1936), cannot be identified as belongiirg with this group. However, the basic similarity of the component species mak-esthis task somewhat sinpler than it might otherwise be. The following attributes seen reasonable: skull not unlike that of the primitive accipitrid, differing in the ossification of the vestibular wal1s, in the detailed relationship of the premaxillary palatal extensions and the medial ossified bar, in the minutia of the cranium and palate, and in the lack of superciliary, s€parately ossified anteropterygoid, and jugal; mandible with a meckelian fenesta; coracoid slin with rounded head and large procoracoid which nearly encloses the triosseal canal, distal end only slightly expanded and without a distinct sternocoracoid process; sternum with two deep notches posteriorlL keel deepest at the anterior end; several thoracic vertebrae fused, reduced number of vertebrae in synsacrum; pelvis broad with rounded contours; tarsometatarsus with inner calcaneal ridge extending nuch of length of bone, leg and foot not powerfully developed; ventral feather tract with strongly separated axilLary FALCONIFORMES r23(323)

and sternal bands, vestigial connection between large axillary division and ventral tract at inner base of thigh, dorsally the prinitive bilateral bands meet only in the caudal region, meta- tegnen developed and the after shaft retained; color pattern much like that of the accipitrid (nodified along similar lines in the evoLution of the group) ; the lores and cere covered with bristles; tarsometatarsus reticuLately scaled. The history of this group can only be conjecture, It can be assumed to have developed from an arboreal line, the foot adapted to perching. Probably always weakly predaceous, this group fed on small organisns killed by nipping with the bill. The feet were used for holding prey, but they lacked the clenching power of the accipitrid. Radiation within the group was nainly a matter of perfection of predation styles with some species tending toward scavenging. This group arose in the New World (South funerica?) to fill a predation niche not occupied, or only partially occupied, by the accipitrids. The most successful genus, Faleo, with several different sizes of species, invaded the 01d World and is now cosnopol itan. 0f the living genera, Mlluago is most prinitive in plunage coloratiori. This plumage feature is shared with Polybonus Lutosus, reraeidea benigora" and Falco nooaeholLandi,ae (and the immatures of many species of Faleo). Mienaetur and Herpetotheree represent o1d, relict species with speciaLtzed color patterns and a conbination of speciaLizations and prirnitive features. Parti- cularly to be noted is the great alteration of the tarsometatarsus in Micraatuy as compared with Henpetotheres or the other falconids. Alteration of this structure is also observed in the Polyborinae or Spiziapteryx.- It is not really possible to identify one genus as the most prirnitive since all of the genera have undergone change with time lfig. 2L0). The difference between Mierastur and Herpetotheres is as great in some details (form of tarsometatarsus) as a comparison of either with FaLeo. Yet, for example, these genera are nore alike and more specialized, in terms of the soleus muscle than either is like laleo. The various criteria suggest the great age of Mi.erastur and Henpetotheres. The large number of species of FaLeo suggest that this radiation occurred relatively recently and was due to some extent at least to the periodic ice ages of the PLeistocene. In the genus Faleo, radiation has involved a general reduction in size. THE PLACEOF THE FALCONIFORMSIN THE CLASSAVES

Introductory Remarks Various sections of this study establish the view that the types found in the order Falconiformes are so dissinilar that their inclusion in a single order is undesireable. The evidence does not prove this conclusion, but linited comparisons in each systen exanined indicates that the degree of difference between these types is in each case at least as great as that between other orders. Within each of the falconiform groups there is a range of 124( 32\) M. JOLLIE

Falco

"Cerchneis"

I'I eracidea"

Microhierax

Po1 ihierax

Neohierax

Spiziapteryx

Polyborus

Phalcoboenus

Daptrius

N{ilvago

Micras tur

Herpetotheres

Fig. 2L0, The phylogeny of the falconids.

variation comparable to that of some orders and nowhefe, in spite of the tonvergent forces which rnight be p_resumed to be in operation, is there conclusive overlapping of features. Where cLrtain features are shared, they are of doubtful value. The question then becomes one of where do these types belong,in !h9 pnyfogeny of birds? Each of the groups can now be considered in ttre light of the broader results of this study.

The Cathartid Type This type is the most distinct of the falconiforms. Its uniqueness does not appear to stem from specializatton; rather, it ieflects a different heritage, a heritage which appears to be shared to a noticeable extent with the procellariiforms, pelecaniforms, ciconiiforms, and charadriiforms. On the basis of its morphological peculiarities, when compared with these types, it should be considered an independent order. This type is represented in the upper Eocene by Neoeathartes grallator. i'tris fosiil appears to be essentially like the living species and shows no greater resemblance to the other falconiform types than they do. The most aberrant member of this groPP is tha Pleistocene species, Tenatornis merriami. This huge bird appears to be greatly .rnodified, away from the basic primitive style as typified by the smaller living species of Cathartes and Coragyps. Its inoaifications are paralleled to some extent in large species of re1-ated orders. FALCONIFORMES t25(325)

The suggestion by Ligon (1967) that the cathartids are a suborder of the ciconiiforms is rejected. I have reviewed each of the areas exanined by Ligon but reached a different conclusion. The possible role of the cathartid as a precursor of the accipi.- trid or sagittariid types will be discussed under the next head- ing.

The Sagittariid Type Many of the renarks concerning the cathartid and its next of kin apply equally well to this group, and any discussion of the origin of the accipitrid from a "ciconiiform" group (Garrod, 1874), or "Order Pelargornithes" Ftirbringer, 1888) involves both types. First let us consider the degree of interrelationship and then the "primitiveness" of position of both groups. The assumption of a relationship between the cathartid and sagittariid types cannot be confirned by any of the norphological features described here. There is a superficial sinilarity in some details of the hind leg musculature, and the form of the tarsometatarsus but nothing more. Whatever features these two types share, as suggested by the description of their archetypes, th-ese are very generalized structures shared probably with the entire aquatic array. Whereas the cathartid shows some evidence of vultuiine specialization, the sagittariid, as compared with the ancestral type, is greatly modified in its head skeleton, pterylosis, and-crane-like form. Diatnopornia eLLioti, known fron seve.ral tarsometatarsi and the distal end of the tibiotarsus (Gai11ard, 1908: Cracraft and Rich, t972) couLd be considered 'iinterrnediate" in forn but the evidence is far too weak to have real neaning. These tarsometatarsi might belong to a number of orders. Their sma1l size suggests a fairly specialized species, well away fron any stem position. The suggestion (supported by the long-legged Neoeathartes) that the primitive falconiform was a long-legged ciconiine does not seem very probable, although the ancestor was probably mofe cursorial, with longer legs, than nany of the living species. Crane-like proportions are a specialization and, therefole, a matter of convbrgence in the several types sharing then (Sagittar- ius, cranes, carianids, some ciconiiforms, flaningos) . Since the cathartid and sagittariid show little agreement' it is not likely that they represent remnants of a single primitive, predatory type, but more likely that they are separate evolutionary lines which have retained some features in common with other, equally distinct and separate lines such as the pfo- cellariiforms, ciconiiforms, pelecaniforns, stercorariids, and possibly- gulls. The cathartid type as an ancestor of the accipitrid seems most unlikely. If the cathartid constitutes an effective vulturine type, which is suggested by its long history, it would seen probable that it could have evolved into some predatory form. Failure to become active predators is best explained as due to occupation of such niches by agressive, well-adapted accipitrids. The contenporary evolution of these two groups is indicated by their structural differences, which must be considered as heritage rather than habitus. ]-26(326) M. JOLLIE

Sagittani-us as a living representative of the prinitive accipitrid stem type, cannot be so easily disnissed. If the head skeleton is indicative of commonancestry, one can assume that the total predatory form (i.e., that of the accipitrid type) had already been fixed before their dichotony. Such a conclusion does not agree with the stork-like body skeleton and the distinctive pterylos is and rnyology. The s tyle of pterylos is (more special -, ized than that of the accipitrid) might easily be the result of the alteration of body form, for it is parallel,ed by changes among ciconiids. Certain features of the skeleton and nyology appear also to be functional dictates of this body forn (agreement with storks, cranes, and Cariama). However, whereas other stork-like types reveal their natural interrelationships-in many details, Sagiltariue does not. Even in terns of the head structure there is doubt as to an accipitrid relationship. This head type could be a functional convergence, 4 situation apparently- p-aialleled by the falconid. The differences betwepn the accipitrid and sagittariid types suggests that divergence occurred at such an earLy stage, that ordinal separation better expresses what must otherwise be anatonicalLy meaningless. The Accipitrid Type If we assume that the accipitrid type shared a conmon ancestry with the sagittariid, it was not until after their divergence thbt the accipitrid differentiated into what has been described here. The constancy of pterylosis' osteology, and nyologyr- neasured in terms of the- aberrant genus Pandion, suggests that this line stems back to the initial radiation of birds. ,Fossil remains of accipitrids date back to the Upper Eocene. These agree well with living forms, and there is no evidence of gradation toward other types (orders) of birds, including Sagittarius. The fossils suggest that m?ny primitive features haire been retained in some of the vulturine genera, particularly Gypohiez.ar. In size the early accipitrids were fail|1-large' firger perhaps than most of the rnodbrn species. Modification in thi; grbup has apparently involved an average decrease in size. Wheri:as thecathartid and sagittariid show some sirnilarity to a group of orders, the accipitrid type does not. The accipitrid is a radical departure in form of skull and tarsometatarsus and in myology, particularly that of the hind limb. The-physical isolatioir of-the- accipitrid type suggests a long independent history. The body plan of the accipitrid does not clearly indicate its ancestry, but one can assune that it came from a cursorial, vulturine type which was contemporaneouswith a second type that eventually gave rise to the cathartiform, procellariiforn, pelecaniforn, ciconiiform, sdgittariiform, sphenisciform, and cftaradriiforrn arrys. This ancestral type dates back at least to the Cretaceous. 'ilre Falconid 'l'ype The possible relationship between the accipitrid and the falconid types is the most difficult to discuss. It nakes no FALCONIFORMES r27(327 )

difference whether the problem is viewed in terms of the two types as derivatives of an intermediate ancestor or as products of a prototype belonging to one or the other of the present types (1 .e.-' ialconid type anceitral to accipitrid or vice versa). The lack of overlap between them does not agree with a theory of common an- centry. For example, if the style of pterylosis'of the accipitrid is adapted for predation, then both types should share it. question then is why are the two types so different? The same The -of type analysis of skuil forn, tarsometatarsus, and rnyological- features reaches this same unalterable conclusion, viz., that the starting points of the two types were dissimilar and that aqy features shared (hooked bi11, general shape of head, and body proportions) are convergent developments and adaptations to a way of 1ife. Sushkin (1905) and others have assumed that Herpetothenes or Micrastur is indeed interrnediate between the accipitrid and fa1- conid types, a conclusion not reached in this study. Aly greater sinilarity between these genera and the accipitrid can be cbn- sidered more as evidence of convergence than relationship, a rnuch less than one rnight expect. degree- of convergence If the accipitrid and falconid types are not related in some direct way, where should one look for their next of kin? The falconid can be based on an hypothetical point-s of origin of the imilarity to owls, parrots, cuckoos and plantain. o'steological - eaters. It is doubtful that the pectoral girdle of the falconid would show convergence to that of such a diverse lot. Rather a more intinate relationship can be assumed between these groups_ than exists with types tacking this kind of pectoral girdle. In contrast, the pectoral girdle of the accipitrid resembles more the water bird type than the arboreal sten of the falconid. The constant conclusion reached is that the falconid is more closely affiliated with other non-falconiform types of birds than the The p?st association of these typ-es is based wholely"ctipitrid. bn their preditory habit and not upon norphological similarities, other than those which are patently adaptive.

Summary and Conclusions The order Falconiformes, as defined by Peters (L931) , Mayr and Amadon (1951) and ltletmore (1-960), has been examined in as much detail as possible in terms of its ptilology, osteology, myology, viscerology, and external features. Conparisons are eiamiiea in terms of demonstrating similarities or lack of sinilarity. The evolutionary trends of features are considered along with ttre idea of each species being a mosaic of features. The conclusion, suggestive or strongly reinforcing, of each of these anatomical accounts has been that the falconiforms is an artificial aggregation of species of four orders: Sagittariiformes, Catharlifoimes, Accipitriformes, and Falconiformes. These orders are separated by morphological gaps as large 85r or_larger than, those separating many other orders. llhereas some of the present orders of birds can confidently be associated into even larger categories, the Accipitrifornes stand alone as the earliest cur- soriil, scavenger-predatory line of birds. The Sagittariifornes 128( 328) M. JOLLIE

and-Cathartifornes appear to be natural parts of a vast array of birds of aquatic or shore habitats whereas the Falconiformes (seneu stpieto) belong with an arboteal array. LITERATURECITED

Aldrovandi, Ulyssis. 1581-1603. 0rnithologie, hoc de avibus historiae (Philosophi ac Medici Bononiensis Historian Naturalem in GyrnnasioBononiensi Profitentis) . (in 12 books, to publisher given). . 1910. DeMusculis Aquilae: Ornithologia lztL7. Frankfurt. (not seen). Al-Hussaini, A. H. 1938. Notes on the anatomy of the Egyptian Kite, Milvus aegyptius (Gmelin) Egyptian Univ. BulL. Fa- culty of Science, No. L6 :1- 16. Alix, Edmond. L874a. Memoire sur 1'osteologie et myologie du Nothura Major. Jour. de ZooI. (Paris), 3: 167-?,L4; (Part II): ?.52-285. . 1874b. Essai sur 1'appareil loconoteur des Oiseaux. ------ETTs. Libraire de 1'Aced6mie de M6decine. Banzhaf, Walter. L929-1930. Die VorderextrenitEt von 0pistho- conus cristatus (Vieil1ot) . Zeits. Wissensch. Bio1. , Abt. A, Morph. und 0eko1. , L6:1L3-233. BarnikoL, A. L952. Korrelationen in der Augestaltung der SchHdelform bei Vdgeln. Morph. Jahrb., 92:373-4L4. --- prrse""l?3:f,;'nff :;':*"ft*n?"3:;-H:3in3t3,l?5:*:3:"n Musophagidae, Galli, Colunbae und Cuculi. ZooL. Jahr., Abt. System. , 81:487-526, . 1953b. Zur Morphologie des Nervus Trigerninus der ---T*oJ e fun t er Be s onde r er Be r i.icks i ch t i gung der Acc i p i t r e s, Cathartidae, Striges und Anserifornes. Zeitsch. Wissensch. ZooL., L57:285-352. Beddard, Frank E. 1.889. On certain points in the anatomy of the Accipitres, with reference to the affinities of Polyboroides. Proc . ZooL. Soc, London, (1889) : 77-82. . 1898. The structure and classification of ffi York. Longnans, Green and Conpany. . 1905. On the nodifications of structure in the syrinx of the Accipitres, with remarks upon other points in the anatomy of that group. Proc. ZooL. Soc. London., (1903): L57- L63. Beecher, William J. 1950. Convergent evolution in the american orioles. Wilson Bull. , 62:51-86. . 195L. Adaptations for food-getting in the reckbirds.amerlcal Auk, 68:41L-440. 195J. A phylogeny of the Oscines. Auk, 7O iZTO- 333. Berger, Andrew J. 1956a. The expansor secundariorum muscle, with special reference to passerine birds. Jour. Morph., 99:L37- 165. . 1.956b. The appendicular nyology of the pygny falcon (PoLihienan semitorquatue) . Am. Mid. Nat., 55:326- 353. FALCONIFORMES r29(329)

. 1957. 0n the anatony and relationships of fregTTupue uarius r Br extinct starling from the Mascarene .Is1ands. 8u11. Amer. Mus. Nat. Hist., 1L3z23l-272, ffig7:l",}'"i;22,::'illI;""'3ili.,'t;,ll;:{};:''of Bignon,- Fanny. 1889a. Contributions a 1'6tude de 1a pneumaticit6 chez 1es oiseaux. Les ce1lules a6rienne cervico-c6phaliques de oiseaux et leurs rapports avec les os de la tete. Men. de la Soc. ZooL. de France (Paris), 2:260.-320. . 1,889b. Note sur Les r6servoirs adriens de @thantesatratus).Conpt.Rend.Hebdonadaires des S6ances et M6rnoires de La Soc. de Biol. (Paris). Ser. 9, L :39- 40. Blanchard, E. L860. Recherches sur les caractBres ost6ologiques des oiseaux, appliqu6es a 1a classification des ces animaux. Preniere partie: Passereaux des ornithologistes. Conpt. Rend. Acad. Sci. (Paris) , 50:47- 50. Boas, J. E. V. L929. Biologisch-anatomische Studien i.iber den Hals der V6ge1. K. Danske Videnskab Selsk. Skr., Naturv. og. Math. Afd., Ser. 9, L:105-222, Bock, W. J. 1963. The cranial evidence for ratite affinities. Proc. L3th Int. Ornithol. Congr., t962:39-54. Boetticher, Hans von. L92,9. Morphologische und plylogenetische Studien iiber die Hornige Fussbekleidung der V6ge1. Jena Zeitsch. Nat., 64:.377-448. Biiker, Hans. 1,929. Flugvern6gen und Kropf bei Opisthocomus cristatus und Stringops habroptilus. Morph. Jahr., 632L52- 207. Boulton, R. 1927. Ptilosis of the house wren (froglodytes aedon aedon) . Auk, 44:387- 4I4 . Brandt, J. F. 1839. Beitrlge zul^ Kenntniss der Naturgeschichte der Viigel nit besonderer Beziehung auf Skeletbau und ver- gleichende Zoologie. (in Sechs Abhand). Mdn. Acad. Imp. Sci. St. Petersburg, Nat. Sci. ser.6,5:vi-L55. . L853. Die Gruppen und Gattungen der Raubv6gel ---ffi5Tands in exomorphischer und craniologischer Beziehung. Jour. f. Orn., 1:1-78-195. Breder, C. M., Jr. 1947. An analysis of the geometry of symmetry with especial reference to the squamation of f i.shes. 8u11. Am. Mus. Nat. Hist., 88:325-4L2. Brodkorb, P. 1963. Catalogue of fossil birds. Part 1. Bull. Fla. State Mus., 7:L79'293. Brown, Leslie, and Dean Amadon. 1968. Eagles, hawks, and falcons of the world. 2 vols. New York, McGraw-Hil1. Burckhardt, Dieter. 1954. Beitrlge zvT embryonalen Pterylose einiger Nesthocker. Rev. Suisse de Zool., 61:551-653. Buri, Rud. O. 1900. Zur Anatomie des Fliigels von Mievopus nelba und einigen anderen Coracornithes, zugleich Beitrag zur Kenntnis der systematischen Stellung der Cypselidae. Jena. z. Naturwiss., 33:361-61-0. Burt, William Henry. L929. Pterylography of certain North Aneri- can woodpeckers. Univ. Calif. Publ , ZooL. , 30:427-442. Chamberlain, Frank Wilbur. 1945. Atlas of avian anatomy. 0steol- ogy-arthrology-myology. Mich. State College Agr. Exp. Station, Mem. Bu11. , 5:l- 47 . 130(330) M. JOLLIE

Chandler, Asa C. 1914; Modifications and adaptations to function in the feathers of Circus hudeoniue. Univ. Caiif. Publ. Zool. , 11: 329--776, . 1916. A study of the structure of feathers, with reference to their taxonomic significance. Univ. Calif. Publ . ZooL.,-L932, L3:243-446. Clrapin, James P. The birds of the Belgian Congo. Part I. BulL. fun. Mus. Nat. Hist,, 65:1-765. Chapnan, Frank Michler. L926. The distribution of bird life in Ecuador. A contribution to a study of the origin of Andean bird life. Bul1. Am. Mus. Nat. Hist., 55:L-783, Chonjakoff, M. 1901. Zur Entwickelungsgeschichte des Sch[de1s einiger Tagraubvogel. Anat. Anz., 19: L35-140. Clark, W. E. LeGros. L955. The fossil evidence for human evolution: an introduction to the study of paleoanthropology. Chicago, Scientistrs Library, Biology and Medicine. Compton, Lawrence. V. 1938. The pterylosis of the Falconiformes with special attention to the taxonomic position of the Osprey. Univ. Calif . Publ . ?ooL., 42:L73-2L2. Condon, H. T. L95L. Variation in the Brown Hawk. Enu, 50:LSZ- L74. Cope, E. D. L877. Report on the extinct Vertebrata obtained in New Mexico by parties of the Expedition of I874. Geol. Surv. W. of 100th Meridian by Geo. M. Wheeler. Vol. 4, Paleontology: L- 370. Cornay, Joseph E., Dr. L844. Apercu d'une nouvelle clasqifica- tion des oiseaux, fond6e sur des considerStions tirdes de lfos palatini . Compt. Rend. Acad. Sci . Paris . , 28:94- 95 . , 1847. Consid6rations g6n6rales sur La cTassTficlilIoi-des oiseaux, fond6e sur la consid6ration de lros palatina anterieur. Rev. ZooL., Paris, (1847)2360-369. Coues, Elliott. 1884. Key to North American birds; containing a concise account of every species of living and fossil bird at present known fron the continent north of the Mexican and United States boundary, inclusive of Greenland. Boston, Estes and Lauriat. . 1890. Handbook of field and general ornithology. ffiofthestructureandc1assificationofbirds. London, MacrniI l an and Co. L td . Cracraft, Joel and Pat Vickers Rich. L972. The systenatics and evolution of the cathartidae in the Old World Tertiary. Condor, 74:272-283. D'A1ton, Dr. 8., d.A., und Df Alton, Dr. E., d.J. (DrAlton und Pander)1838. Die Skelete der Raubv6gel. Bonn, Edward Weber. deBeer, G. R. L937. Development of the vertebrate skull. Oxford. Clarendon Press. Delacour , J . , et. F. Edmund-81anc. 1933- 1934. Monographie des Veuves (Revision des generes Euplectes et Vidua). LrOiseaux et La Revue Francaise d'Ornith., 5:519-562; 687-726: 4:52-11.0. Dwight, J. Jr. 1900a. The sequence of plurnages and nolts of the passerine birds of New York. Ann. New York Acad. Sci. , 13:73-360, FALCONIFORMES 131( 331)

----(q ; r1?l:?;s"lh:"1"1:"l:"i1"ffiil::t?:li:31,'i;liit;l"" Engelnann, Fritz. L928. Die Raubvdgel Europas: Naturgeschichte, Kulturgeschichte und Falknerei. Neudamn,J. Neumann. Engels, Williarn L. 1940. Structural adaptations in thrashers (Mirnidae: genus Toxostona) with commentson interspecific relationships. Univ. Calif. Publ . ZooL., 42:341-400. Erdmann, K. 1940. Zur Entwicklungeschichte der Knocken im Sch?idel des Huhnes bis zum Zeitpunkt des Ausschliipfens aus dernEi. Zeits. Morphol. Oeko1., 36:315-400. Eyton, T. C. 1867-1875. Osteologia Aviurn; or, a sketch of the osteology of birds. London, Williams and Norgate. Fisher, Harvey I. 1939. Pterylosis of the Black Vulture. Auk, 56:407-410. . 1,943. The pterylos is of the King Vulture. Condor, 45:69-73. . L944. The sku11s of the cathartid vultures Condor , 46: 272- 296. . 1945. Locomotion in the fossil vulture Teratonnis . Am. Mid. Nat., 332725-742. 1946. Adaptations and comparative anatomy of the locomotor apparatus of the New World vultures. Am. Mid. Nat. , 35: 545-727. t947. The skeletons of recent and fossil Gymnogyps. Pacific Sci. , I:227-236. . 1955. Avian anatony, 7925-1950, and sone suggttted problems. rn Wolfson, Albert (ed.). Recent advances in avian biology. Urbana, Univ. Illinois Press. pp. 57- L04. and Donald C. Goodman. l-955. The myology of the Whooping Crane , Gnus amerieana,. I11 . Biol . lrlonogr. , 24:I- L24 . Fit zinger, Leopold Jos . 1856 . tiber das Sys tern und die Charak- teristik der Natijrlichen Farnilien, der Viigel. Sitzungber. Kais. Akad. d. Wiss. Math.-Nat. (Vienna), 2t:277-3I8. Forbes, W. A. 1882. Report on the anatomy of the petrels (rubinares), collected during the voyage of H. M. S. Challenger. Report gn the scientific results of the Voyage of H. M. S. Challenger during the years 1873-I876. 4:I-64. . theclassirication or --'rrdr. rbl:lo;",:ttitriti3_iiol" Friedmann, Herbert. 1"930. Birds collected by the Childs Frick Expedition to Ethiopia and Kenya Colony. Bul1 . U. S. Nat. Mus. 153: Part I : Non Passeres . xiii- 516. . 1950. The birds of North and Middle America Part XI. Bul1. U. S. Nat. Mus.50:xii-793. Fiirbringer, Max. 1888. Untersuchungen zur Morphologie und Systematik der Viigel zugleich ein beitrag zur Anatomie der Stutz und Bewegungsorgane. K. Zoo1. Genoots. Bidr. tot de Dierkunde (Amsterdam, I. J. Van Holkema.), 15 (2 parts). . t902. Zur Vergleichenden Anatomie des BTustsElulterapparates und der Schultermuskeln. Jena. Zeit. Naturw., 36:289-736. 132( 332) M. JOLLIE

Gadow, Hans. 1879. Ve,rsuch einer Vergleichenden Anatonie des Verdauungssystemes der Viigel.. Jena . Zeitschr. Naturw., 15:92-L7L; 339-405. . L889. 0n the taxononic value of the intestinal ---- convoTutions in birds. Proc. LooL. Soc. London, 32:303-316. 3::' rnl3f: i", .f3,'il 333 ;,3il;'T i' !;' 51?i33t" llt'". Wort uia gifa. Sechster Band. IV Abt., YoL. 2, Systenatische --rrn;""'Thei1. Leipzig und Heidelberg, C. F. Winterrsche. and-Emil Selenka. L891. Viigel , rn Bronns, Dr. H. G.' --TLassen und 0rdnungen des Thier-Reichs, wissenschaf tlich dargestellt in l|hrt und Blld. sechster Band. IV Abt. , _ vol. L, Anatonischer Theil. Leipzig und Heidelberg, c. F. Winter t sche. Gai11ard, C. 1908. Les oiseaux des phosphori!"? du_Quercy._ ' Ann. ae I'Universitd de Lyon, Nouv. ser. Sci. Med., 23:L-,128. Garrod, Alfred H. L873. Part I. On certain muscles in the thigh oi Uirds and their value in classification. Proc. ZooL. Soc. London,'. (1873)-L874'. :626-644. Part II. on certain nuscles in the -i-- tTigilof-5-irds and their value in classification. Proc. Zool. Soc. London, (1874) :LLL-L23, der Viigel. Gegenbaur,- Carl. t873. Ueber die Nasenmuscheln Jena, Zeitschr. Naturw. 7:L-2t. George, J. C. and A. J. Berger. 1966. Avian Myology. New York Academic Press. Gerber, A. 1939. Die embryonale und post-ernbryo+qle_Pterylose d6r Alectoromorphae. Rev. Suisse ZooL,,, 46:16t-324. Gilbert, Perry W. fbSg. The avian lung and air-sac systen. Auk, 56:57-63. Glenny, Fred H. 1941. Presence of the liganentun botalli in ttre golden eagle, red- tailed hawk, and the commonPigeon. Ohio Jour. Sci., 41:46-49. . L955. Modifications of Pattern in the aortic @em of birds and their phylogenetic significance. Proc. U. S. Nat. Mus., 104:525-62L. GoodchiLd, J. G. 1886. Observations on the_disposition of the cubiial coverts in birds. Proc. Zool. Soc. London, (1886): 184- 203. Goodrich, Edwin S. 1930. Studies on the structure and de- vel6pment of vertebrates. London, Macrnillan and Co. Ltd. Grass6, Pierre-P. (Ed..) 1950. Trait6 de Zoologie:_ Anatornie, Syst6matique, tiiotogie. Tome XV. 0iseaux. Libraires de liAcad6mie de M6decine. Paris, Masson et Cie. monotremes and the palinsest Gregory,- William King . L947. The theory. Bul1. An. Ir{us. Nat. Hist., 88:6-52. Griscon, Ludlow. L932. The distribution of bird life in Guatemala; a contribution to a study of the origin of central American bird 1ife. 8u11. Am. Mus. Nat. Hist.' 64:1-425. Hamilton, A. 1892. 0n the fissures and caves at the Castle Rocics, Southland (New Zealand) : with a description of the remains of the existing and extinct brrds found in then. Trans . New ZeaL. Inst. , 25 :88- 106. FALCONIFORMES 133(333)

. 1893. Results of a further exploration of the bones fissu::es at the Castle Rocks, Southland. Trans. New ZeaL. Inst., 26:226-229. Hennig, WiL1i. 1966. Phylogenetic systematics (translated by D. Dright Davis and Rainer Zangerl). Urbana, Univ. I11. Press. Heilmann, Gerhard. L926. The origin of birds. New York. D. Appleton and Co. Hofer, Helmut. 1945. Untersuchungen iiber den Bau des Vogel- schldels, besonders tiber den der Spechte und Steisshiihner. ZooL. Jahrb. (Abt. Anat.), 69:1-158. . L949. Die Gaumenliickender V6ge1. Acta . ZooL.

. L955. Neuere Untersuchungen zur Kopfnorphologie --TeJT'ge1 . rn Portmannr' Adolf and Ernst Sutter (ed. ) . Acta XI Congressus Internationalis Ornithologici Basel. Basel and Stuttgart, Birkhduser. pp. L04-L37. Holngren, Ni1s. L953. 0n the origin of the tetrapod limb. Acta. ZooL. (Stockholn) , L4:1-85-295. . 1939. Contribution to the question of the otfgln of the tetrapod limb. Acta. ZooL. (Stockholn), 20:89-I24. . 1955. Studies on the phylogeny of birds. ffir. (stockholm), 36:42-328. Howard, Hildegarde. L929, The avifauna of Eneryville shell mound. Univ. Ca1if. Pub1. Zool. , 32:301-394. 1932. Eagles and eagle-1ike vultures of the Pleistocene of Rancho La Brea. Contributions to Palaeontology, Carn. Inst. Wash. Pub1. no. 429zl-ii-82. 1935. A new species of eagle from a quaternary cave deposit in eastern Nevada. Condor, 37: 206-209 . L937-1938. Aves. Fortschritte der Paliiontologie. 2:309 - 327,. 1938. The Rancho La Brea Caracara: a new specr-es Carn. Inst. Wash. Publ . no. 487:2L7 -240. L944. A Miocene hawk from California. eoilI-r 46:736-237. . L946. A review of the Pleistocene birds oE-Tossif take, Oregon. rn Furlong, E. L., E. Raymond Hall Lore Rose David, Robert E. Wallace. Fossil vertebrates Paleontolosv'from western North America and Mexico. Contributions to .t'il;r1"'i;XilS;.,llol;rt3; Sllil-l;l3l; a queFtf on-fn-Taxonomy. Auk. , 64:287 -29L. . L952. The prehistoric avifauna of Smith efteT-eete; Nevada, with a description of a new gigantic raptor. Bul1 . So. Calif. Acad. Sci., 51:50-54. Hudson, George Elford. 1937. Studies on the muscles of the pelvic appendage in birds. An. Mid. Nat., 18:1-108. . 1948. Studies on the muscles of the pelvrc appendage in birds II. The heterogeneous order Falconi formes . Am. Mid. Nat. , 39 : L02- L?7 . 134(334) M. JOLLIE

and P. J. LanziLlotti. 1955. Gross anatorny of the wing nuscl-es in the Fanily Corvidae. An. Mid. Nat. , 55 zL-44. Hutt, F. B. and Ball, Lelah. L938. Number of feathers and body size in passerine birds. Auk., 55:651-657. Huxley, Thonai H. 1867. On the classification of birds; and on the taxonomic value of the nodifications of certain of the cranial bones observable in that class. Proc. ZooL. Soc. London, (1867) :415- 472. Hyman, Libbie Henrietta. L942. Conparative vertebrate anatony. Chicago. Univ. Chicago Press. Ihle, J. E. W. L947. Leerboek der vergelijkende ontleedkunde van de Vertebraten. 3rd Ed., 2 vols., N. V. A. Oosthoekrs Uitgevers Mij Ingran, Collingwood. 1920. .A contribution to the study of nestling birds . Ibis , ser. 11, 2 :856- 880. Jol1ie, Malcolrn. L947. Plunage changes in the Golden Eagle. Auk, 64:549- 576. . 1957a. The head skeleton of the chicken and iernarks on the anatony of this region in other birds. Jour' Morph' ' tlghltn;XiX; anatonyand the anatonist. Condor, 59 t394-397, . 1-958. Commentson the phylogeny of the skull

. 1961. Article "Beakr'. rn Peter Gray (ed), the Encyclopediaia of the Biological Sciences. New York, Reinhold. pp.98-99. . L962. Chordate Morphology. New York, Reinhold. ffiighton.L925.Theparabasa1cana1andnervefora- mina and canals in the bird skull. Jour. Roy. Soc. New South Wales (Sydney), 59 : 108-123. Kripp,- Diminik V. 1933. Der Oberschnabel-Mechanismusder Viigel. (Nach den Methodeh der graphischen Statik bearbeitet.) Morph. Jahrb. , 7L:469- 544. Lakjer, Tage. L926. Studien i.iber die Trigeminus-Versorgte Kaunuskulatur der Sauropsiden. Kopenhagen. C. A. Reitzel. Lanbrecht, Kalman. 1935. Handbuch der Palaeornithologie. Berlin, Gebruder Borntraeger. Lang, Charlotte. 1956. Das Craniurn der Ratiten mit besonderer Beriicksichtigung von Struthio camelus. Zeit. f . Wissen. Zoo1. , 159:165-224 LeopoLd, A. Starker. L943. The moLts of young wild and donestic turkeys. Condor, 45: 133-145, Lesson, Rene Primevere. 1830-L83L, Trait6 dtornithologie, ou Tableau m6thodique des ordres, sous-ordres, fami1les, tribus, genresr,sous-genres et races droiseaux. 2 vols. Paris and Strasbourg, F. G. Levrault. Bruxelles, Librairie parisienne. Ligon, J. David. 1,967. Relationships of the cathartid vultures. Occ. Papers Mus. ZooL., Univ. Mich. No. 6512L-26. Lillie, Frank R. L942. On the development of feathers. Biol. Rev. , 17:247-266. Lindsdale, Jean M. t928. Variations in the Fox Sparrow (Passen- eLLq iliaea) with reference to natural history and osteology. Univ. Calif. Publ . TooL., 30225L-392. FALCONlFORMES 135( 335)

Lowe, Percy R. L931a. 0n the relations of the Gruinorphae_to the Charadriinorphae and Rallinorphae, with special reference to the taxononic position of Rostratulidae, Jacanidae, orderand Burhinidae,(':t'?;:l;:nnffi) (0edicnenidae Olim) ; with a suggested new ;""l3ii;"i"1;"13;li'?l;tl*;u",,,, ITe1natomorphae), with special reference to the families, subfanilies, and genera within the suborders Linicolae, Grui-Limicolae, and Lari-Lirnicolae. Ibis r ser. 13, l": 712-77L. Lucas, Alfred M. and Peter R. Stettenheim. L972. Avian Ana- tomy: Integument Parts I-II, Agriculture Handbook 362, Washington, U. S. Gov. Print. Office. Lucas, Frederic A. 1882. Notes on the os proninens. Bu|l. Nuttall Ornith. c1ub, 7: 86- 89. Macalister, Alexander. 1864-1866. 0n the anatomy of the 0strich. (Struthio camelus). Proc. Roy. Irish Acad., 9:L'24. Mcnowell, Sam. 1948. The bony palate of birds. Part I: The Paleognathae. Auk, 65: 520-549. Mclean, Franklin C. and Urist, Marshall R. L955. Bone. An introduction to the physiology of skeletal tissue. Chicago, Univ. Chicago Press. ilber den Bau des Kndchernen Magnus,- Hugo. fgZf. Untersuchungel V6gelkopfes. Zeits. wiss. Zoo1. , 2l:L- 108. Maillard, Joseph. 1948. Recherches embryologiques sur Catharacta skua Briinn (pterylose et ossification). Rev. Suisse ZooI., 55 (Supplement 1-):1-LL4. Marinelli, Withetrn. L946. Kranium und Visceralskelett. In Bolk, G6ppert, Kallius, und Lubosch, Handbuch der Vergieich-enden Anatomie der Wirbeltiere. Vo1. 4. Berlin and-Wien, Urban and Schwarzenberg. pp. 207'ZSZ. Ntarshall, Margaret E. 1905. A study of the anatomy of Phalaenop- tilus , Ridgway. Proc. Am. Philos . Soc. , 44 2L3--240. May, Frankiin H. 1945. Ptilology-'a proposed name for the general study of the plurnage of birds. Auk, 62:308. Mayr, Ernst and Alnadon, Dean. 1951. A classification of recent birds. Am. Mus. Nov., no. L496zL-42. Menzbier, M. A. 1916. Faune de La Russie. Oiseaux (Aves). Vo1. 6. Falconiformes. Pt. I. Petrograd, Acad. Sci. USSR. Merrem, Blasius. L781. Von den Muskeln des Weisskopfiggt- Ailers: Vermischte Abhandlungen aus der Thiergeschichte. Giittigen, V. Vossiegel, pp. 110-163. Meyer, A. L. 1879-1897. Abbildungen von Vogels\9letten ' der herausgegeben- mit Unterstiitzung der Generaldirektion Kiinigl . Sammlungenf iir Kunst und IVissenschaft in Dresderr. 2 vols. Berlin, R. Friedldnder und Sohn. Mi11er, Alden H. t937. Structural modifications in the Hawaiian Goose (Nesoehen sanduieensis). A study in adaptive evolution. Univ. Calif. Publ. ZooL., 42:1-80. . 1944. An avifauna from the l,ower Miocene of South Dakota. Univ. Calif. Pub1., 8u11. Dept. Geo1. Sci. , 27:85- 100. and Harvey I. Fisher. 1938. The pterylosis of the Cal i fornia Condor. Condor, 40:248- 256. r36(336) M. JOLLIE

Mi11.er, Loye Hol.nes. 1909. Teratornis, a new avian genus from , ' Rancho La Brea. Univ. Calif. Pub1. , 8u11. Dept. Geol. , 5: 305.3L7. 19L1. A series of eagle tarsi frorn the ffiRancho La Brea. Univ. Calif. Pub1., Bul1. Dept. Geol. , 6:305- 316. . LgLz. Contributions to avian palaeontology ffiic coast of North Anerica. Univ. Calif . Pub1. Bul1. Dept. Geo1. 7:61-115. . 1925. V. The birds of Rancho La Brea. Carn. f nst. ltlash. Publ. no. 349:63-L06. , L927. The falcons of the McKittrick Pleistocene. Condor, 29: l-50-152. . 195L. Bird remains from the Kern River Pliocene of California. Condor, 33:70-72. . L937. Skeletal studies of the tropical hawk Hanpague. Condor, 59: ZLg-22L. and Hildegarde Howard. 1938. The status of fh-e exfinct condor-Like birds of the Rancho La Brea Pleistocene. Univ. Ca1if. Los Angeles Pub1., Bio1. Sci., 1:169- L76. Miller, W. DeW. 1915. Notes of ptilosis, with special reference to the feathering of the wing. Bull-. An. Mus. Nat. Hist. , 34:129-L40. . 1924a. Further notes on ptilosis. Bull. Am, ffi Hist., 5o:305-33L. . 1924b. Variations in the structure of the ffit and their taxclnonic value. fun. Mus. Nov., no. L40zL-7 . Milne-Edwards, A. 1867-187L. RechErches anatoniques et paL6ontologiques pour servir a lrhistoire des oiseaux fossiles de la France , Z vols. Paris, Victor Masson et fils. and ALfred Grandidier. 1879-188L. Histoire @ature11eetPo1itiquedeMadagascar.Vo1.xII. Histoire Naturelle deE 0iseaux. 2 vo1s. Paris, 1'Imprimerie Nationale. Mitchell, Peter Chalners. 1896. 0n the intestinal tract of birds. Proc. ZooL. Soc. London, (1896):136-l-59. nli3';",31.1H ll:";iil;:,:li;: :f zoological characters. Trans. Linn. Soc. London, ser . 2. , 8:173-275, Mivart, St. George Jackson. 1873. Lessons in elenentary anatomy. London, Macmillan and Cornpany. Morse, Edward S. L872. On the tarsus and carpus of birds. Ann. Lyceun Nat. Hist., (New York), L0:3-22. Newton, Alfred and Hans Gadow. L893-1896. A dictionary of birds. London, A. and' C. Black. Nitzsch, Christian L. 1811. 0steologische Beitrage zur Naturgeschichte der Vd,ge1. Leipzig. Carl Heinrich. (edited by C. Hermann Burmeister). 1840 @lographie. Halle, Eduard Anton. FALCONIF'ORMES 137(337 )

(edited by C. Giebel). 1863. Zur Anatonie We aus Chr. L. Nitzschrs hand- schriftlichen NachLass mitgetheilt. Zeits. Ges. Naturw. , 2LtI31-L40. (edited by C. Giebel). 1866. Zur Anatomie des Lannergelers. Zeits. Ges. Naturw. , 27:t49- L58. (translated by Phillip Lutley Sclater) . PterYlograPhY. London, for the Roy. Soc. by Robert Hardwicke. Oliver, W. R. B. 1930. New Zealand birds. Wellington, Fine , Arts, Ltd. 1-945 . Avian Evo1ution in New ZeaLand and ffi Emu 45:5s-77; L19-L52, Owen, Richard. 1849. On the anatomy of the Apteryx australis, Shaw. Part II. (Myology) . Trans. ZooL. Soc. London, 32277-301. . 1866. Conparative anatomy and physiology of

Parker, T. Jeffery and Willian A. Haswell. (Revised by C. -effi'tes.Vo1.II.London,LongmansGreenandConpany.Forster-Cooper). 1940. A textbook of zoology. 6th Ed. Vol. II. New York, Macmillan and Company. Parker, Willian Kitchen. 1866. On the structure of the skull iil the ostrich tribe. Phil. Trans. Roy. Soc. London, 156: L13-185. . 1873. 0n the development of the skull in the tiE and sparrow hawk. Mon. Micro. Jour., 9:6-11., 45- 50. . L879. On the structure and development of the birds' Trans. Linn. Soc. London., ser. 2,L: 99- 1.54. -kul1.. 1890. 0n the morphology of the duck and --the' auk tri bes . eunningham Memoirs. Roy. Irish Acad. , no. 6: iv:t32. Parrington, F. R. and T. S. ltrestoll ' 1941' on the evolution of the nannalian palate. Phil. Trans. Roy. Soc. London. B' 230 :305- 355. Perrin, J. B. L875. 0n the nyology of Opisthocomus cristatus. Trans , ZooL . Soc. London, 9 : 353- 370 . Peters, Jaftes Lee. 1931. Check-list of birds of the world. Vo1. 1. Cambridge, Harvard Univ. Press. Portenko, L. L929. Uber den taxononischen Wert der Formen der pal6arktischen Bussarde. 8u11. 1'Acad. Sci. lrURRS. ' Phys. -Math. , ser. 7 , 7:623-652; 8:707 -7L6. Portmann, Adolf. 1955. Die postembryonale Entwicklung der V6gel als Evolutionsproblem. Acta XI Congressus Internationalis Ornthologici, Basel, May 29 - June 5, 1954. Basel and Stuttgart, Birkhbuser. pp. 138- 151-. Pycraft, W. P. 1900-1903. Some point.s on the of the ' norphology palate of the Neognathae. Jour. Linn. Soc. London, 28:343- 357 . . L902. Contributions to the osteology of birds. Part V. Falconiformes. Proc. ZooI. Soc. London, (1902): 277-320. 138(338) M. JOLLIE

. 1903. 0n the pterylography of Photodilus. Ibis. , s,e?.T;3:56-48. A history of birds. London, Methuen and ---ffi-fany. . L910. Rand, A. L. L942a. Birds of the L936'37 New Guinea Expedition. ResuLts of the Archbold Expeditions, 42. Bull. An. Mus. Nat. Hist . , 79z 289- 566 . . L942b. Birds of the 1938-1939 New Guinea Expedition. ------TesuTts of the Archbold Expeditions, 43. 8u11. Am. l'lus. Nat. Hist., 79:425-516. Richardson, F. L942. Adaptive nodifications for tree-trunk foraging in birds. Univ. Calif. Publ. TooL., 46:3L7-368. Riddle, O-cai. 1908a. A study of fundamental bars in feathers. Biol. Bul1 . , 12:165'L7 4. . 1908b. The genesis of fault bars in feathers and ---Jtre cause of alterations of light and dark fundanentaL bars. Biol. Bul1. , L42328-370. collection Ridgway,- Robert, L874. Catalogue of the ornithological oi the Boston Society of Natural History. Part II. Fa1- conidae. Proc. Bost. Soc. Nat. Hist., L6143-72. . 1875. Outlines of the natural arrangenent of The mTconidae. 8u11. U. S. Geol. and Geogr. Sur. of the Territories. F. V. Hayden, U. S. Geologist in Charge. Vo1. 1 (second Series, L874-1875) Bu11. 4:2ZS'?'3L, Robinson, H. W. L924. Haliaeetus' with feathered tarsi. Bull. U, S. Nat. Mus., Ser. 1L, 62824. Saville, D. B. 0. L957. Adaptive evolution in the avian wing. Evolution, Lt: ZLZ'224. Schinz, Hans R. and Ranier Zangerl. L937. Beitrbge zur Osteo- genese des Knockensystems bein Haushuhn, bei der Haustaube und bein Haubensteissfuss. Denkschr. Schweiz. Ges. Natur- wiss., 72:vt, LL7'L64. Schidler, Eiler Lehn. 1931-. Vol. III. Dannarks Fugle nied Henblik paa de i Grdnland, paa Faerderne og i Kongeriget '.'Iceland lorekommendeArter. Rovfugle (Falconiformes). K/benhavn, Gyldendalske Boghandel, Nordis Forlag. Sharpe,-of R. Bowdler. 1899. A hand-list of the genera and species birds. Vo1. I. London, British Museum. Shelford, R. 1900. 0n the pterylosis of the embryo and nestlings of Centropus sinensis. Ibis, 6:654-667 . Schufetrdt, R. W. 1878. Osteology of the Cathartidae. ltlashington, U. S. Geol. Survey , LZt}. Annual Report, PartI (1883):727 -786. . 188L. 0n the ossicle of the antibrachiun as ffiome of the North Ane,rican Falconidae. 8u11. Nuttall Ornith. Club. , 6 :197-203. . 1889. Notes on the anatomy of Speotyto cuni- ilmnan-ypogaea. Jour. Morph., 3 : 115-125. guide to the York, Mac- ni1lan and Conpany. 1891a. A peculiar character referable to the base of the sku1l in Pandion. Auk, 8:236'237. FALCONTFORMES 139( 339)

. 1891b. Some comparative osteological notes on ffierican kites . Ib is , ser. 6 , 3:228'232 . . 1909. Osteology of birds. Albany, 8u11. N. Y. ffi no. L50:1-381. . 191"9. Osteological and other notes on the notrey-etfing eagle of the Philippines, Pithecophaga jefferyi (Grant). Philipp. J. Sci., 1-5:31-58. . L922. On the skeleton of the wedge-tailed @aetus audax Latham). Emu, 2L2295-306. Sibley,-Charles G. and Jon E. Ahlquist. L97?,. A comparative_ 3tudy of the egg white protein of non-passerine birds. BulI. 39. Peabody Mus. Nat. Hist. Yale Univ. Simpson, George Gaylord. 1946. Fossil penguins. 8u11. Am. Mus. Nat. Hist. , 87:1- 33. Snith, Hobart M. t947. Clas.sif ication of bone. Turtox News, 25 z234- 236 . Stanley, A1lan J. L937. Sexual dinorphisn in North Anerican hbwks. 1. Sex organs. Jour. Morph., 6L:32I-339. and Emil Witschi. 1940. Gern cel1 nigration in-ffia-ffi to asynmetry in the sex glands of hawks. Anat. Rec. , 76:329-342. Starck, Dietrich. 1940. Beobachtungen an der Trigeminusmus-- kulatur der Nashornvcigel nebst Benerkungen iiber einige Besonderheiten des Vdgelsch[dels and i.iber die Kiefermuskulatur im allgemeinen. Morph. Jahrb . , 84 :585 . and Albert Barnikol. 1954. Beitrige zur Morpfofugfe der Trigeminusmuskulatur der V6gel (besonders der Accipitres, Cathartidae, Striges und Anseres). Morph. Jahrb. , 94: L-64. Steinbacher, G. 1935. Funktionell-anatomische Untersuchungen an V6g6lfiissen mit Wendezehen und Ruckzehen. Jour. f . Orn. , 8322L4-282. Steiner, Hans. 1918. Das Problem der Diastataxie der V6ielfliigels. Jena. Zeits. Naturw., 55 : 2ZL-496. Stewartl Robeit E. t95?,. Molting of northern yellow-throat in southern Michigan. Auk, 69:50-59. Stieve, H. I9?4. Beobachtungen iiber den rechten Eierstock und d6n rechten Legdan des Hiihnerhabichts (Falco palunbarius K1.) und einiger anderer Raubv6gel. Morph. Jahrb., 54:137-L56. M. 1035. Colymbus" Hesperonnis" Podieeps-: Stolpe,-V6rgleich -ein ihrer hinteren Extremitlt. Jour. f. Orn., 83: 115- L28. Storer, Robert W. L97L. Classification of birds. fn Farner, Donald S. and James R. King (eds.), Avian Biology. Vol. 1. New York, Academic press. PP, 1-18. Stresenann, Ilrwin. 1927 -1934. Handbuch der Zoologie : eine Naturgeschichte der Stbmme des Tierreiches. 7 (part II) Sauropsida: Aves. Berlin und Leipzig, Walter de Gruyter and Co. and V. Stresemann. 1966. Die Mauser der Vdge1. Jour. f .-Gnith. , L07 (Sonderheft) :1'-445. Stresenann, V. and E. Stresemann. 1960. Die Handschwingen- mauser der Tagraubv6gel. Jour. f. Ornith . 10L:373- 403. 140(340) M. JOTLIE

Strong, Reuben Myron'. L902. The developnent of the definitive feather. Bull. Mus. Conp. ZooL., 40:147-185. Sushkin, P. 1-899a. I . Beitrbg:b zur Classification der Tagraubvcigel nit Zugrundelegung der osteologischen Merkmale. Zoo1. Anz., 2?:500-518. --- senilre,iSinf;""#:"flfn"l,:fi+: flffi.u35:1'lfilf "r".i;", . Moscou, L6:L-163, 1900. II. Systematische Ergebnisse osteologischer IFntersuchungen einiger Tagraubvtigel. ZooL. Anz., 23:269-277. . 1905. Zur Morphologie des Vogelskelets. Ver- gleichende Osteologie der Nornalen Tagraubvtigel (Accipitres) und die Fragen der Classification. Teil I. Grundeinteilung der Accipitres. Teil- II. Falken und ihre nHchsten Verwand- ten. Nouv. Mem. Soc. Imp. Natural. Moscou, 1-6: iv-247 . Swann, H. Kirlce (last part edited by Alexander Wetnore) . L913- 1945. A monograph of the birds of prey. 2 Vols. London, Wheldon and Wesley Sylvester-Bradley, P. B. (ed.) 1956. The species concept in palaeontology. Syst. Assoc. Pub1. No.2, Technau, Gert. L936. Die Nasendriise der V6ge1. Zugleich ein Beitrdge zur Morphologie der Nasenh6hle. Jour. f. Orn., 86:5tL-6L7 , Vanden Berge, James C. t970. A conparative study of the appendicular nusculature of the order Ciconiiformes. An. Mid. Nat. , 84: ?,89- 364. Verheyen, R. 1955. Exploration de Parc National de 1'Upenba. Mission G. F. DeWitte (1946-1949), Fasc. 19, 0iseaux. Brussels, Inst. Parcs. Nat. Congo Be1ge. Waternan, Allyn J. and Eddy, H. P. L942. Comparative study of the air sac systen of birds. Anat. Rec., 84:558. Watson, Morrison. 1883. Report on the anatorny of the Spheniscidae collected during the voyage of H. M. S. Challen- ger. ZooL. Chal1. Exp., Part XVIII:L-244. Wetherbee, David Kenneth. 1957. Natal plunages and downy pteryloses of passerine birds of North Anerica. Bul1. Amer. Mus. Nat. Hist. , 1L3: 539- 436. Wetmore, Alexander. L923. Avian fossils from the Miocene and Pliocene of Nebraska. Bul1. Am. Mus. Nat. Hist., 48:483-507. . L926, Description of a fossil hawk from t-he Miocene of Nebraska. Ann. Carn. Mus., 16:403-408. . 1927. Fossil birds fron the Oligocene of edlora-do. trroc. Colo . Mus. Nat . His t. , 7 i3-LS . . 1933. An Oligocene eagle from Wyoning. ffill. Publ. no. 3227. Misc. 87:1-19. 1936. Two new species of hawks from the ffiraska. Proc. U, S. Nat. Mus. , 84: 73-78. . 1937. Bird remains from cave deposits on Great Exuma Island in the Bahamas. Bull. Mus. Conp. ZooI 80:427-44I . L940. A check-list of the fossil birds of North America. Smithsonian Misc. Col1. (Pub1. no. 3587) , 99:L-81. FALCONIFORMES 141( 34r;

. L943. Two more fossil hawks fron the Miocene oE-Tebraska. Condor, 45 : 229- 23L. . 1944. A new terrestrial vulture fron the Fppei Eodene deposits of Wyoning. Ann. Carn. Mus., 30:57-69. . 1950. A correction in the generic nane for Eoeathav,tes gnallator. Auk, 67 :?35. . L95La. A revised classif,ication for the ffior1d.SnithsonianMisc.Co11.(Pub1.no.4057)' LLTzL-2L. . 1951b. Presidential Address. Recent ad- aiffins to our knowledge of prehistoric birds, 1953-1949. In Proceedings of the Xth International Ornithological Congress, Uppsala, June 1950. UppsaLa-Stockholrn, Almquist Q Wiksells Boktryckeri pp. 5L-74, rrnr,--RecenirlSii;, if i3lH'3fli"rii tllil"" . Press' pp' -filfilt: 1o-i3;u. A check-list of the fossir and pre- histoTfc-Ef-rds of North Anerica and the l{est Indies. Snithsonian Misc. Co11. , L51": L- 105. . 1960. A classification for the birds of the ffisonian Misc. Coll., 139:1-37. and E. C. Case, L934. A new fossil hawk frqn fn-e-Ollgocene beds of South Dakota. Contr. Mus. Pal., Univ. NIich., 4:L29-L32. Witherby, H. E. (ed.) L927. A practical handbook of British Birds. YoI. 2. London, Witherby and Conpany

Further Acknowledgments

I want to thank Jane K. Glaser for her careful photo copying of the illustrations. t42(342) M. JOLLTE

For legend of Ftg. 54, see page 80 - (p. L80 of Volure 2) '

FRETIfl(L|-A NASAL c{Pst|-EossrFtcATtotil TiASALSEFTTUM PREilAXLL.I