SUMMARY After further preparation and study, the early Eocene fossil Palaeophasianus meleagroides Shufeldt is found not to be a member of the Cracidae but is instead representative of the Aramidae.

N 1913 Shufeldt described a new fossil bird, Pala1/2ophasianus mdeagroides, from the early Eocene of Wyoming. Shufeldt considereal the relationships of Palaeophasianus to be within the Galliformes, more particularly with the Tetraonidae and Meleagrididae. On the basis of Shufeldt's published figures Brodkorb (1964:303) placed Palaeophasianus in the subfamily Cracinae of the Cracidae with the remark "The shapes of the cotylae and the proximal inner margin of the shaft [of the proximal end of the tarsometatarsus] are reminiscent of Penelope, although the large size recalls Crax." Brodkorb also suggested that since the type was imbedded in matrix, more preparation of the fossil would be necessary before its relationships could be determined with any certainty. During the course of other paleontological work on Oligocene birds I had occasion to examine the type of Palaeophasianus. Through the courtesy of Dr. Malcolm C. McKenna of the Department of Vertebrate Paleontology of the American Museum of Natural History the type was further prepared, thus allowing a more complete study of the fossil. Subsequent examination has revealed that the relationships of Palaeophasianus are not with the galliforms but with the gruiform birds of the family Aramidae. The fossil appears to represent a heretofore unrecognized genus and species of that family. Because the fossil was covered with matrix, Shufeldt's description was incomplete; therefore, it is necessary to redescribe the type material (see Fig. 1). MATERIAL American Museum of Natural History No. 5128, Department of Vertebrate Paleontology; the distal end of left tibiotarsus, proximal and distal ends of left tarsometatarsus, and seven or eight broken pieces of one or more long bones; collected by the American Museum expedition of 1910, Willwood Formation, Elk Creek, east of Dry Camp 2, Bighorn Basin, N. W. Wyoming; age: early Eocene (Gray Bull fauna). Tarsometatarsus.The proximal end of the tarsometatarsus is sinilar to the living .dramus guarauna, but with (1) the anterior metatarsal groove deeper (may be due, in part, to crushing); (2) intercotylar prominehce more horizontal, not projecting upward as much (when viewed from the side); (3) intercotylar prominence well developed but less well defined; (4) internal cotyla larger than external cotyla; (5) intercotylar area more elevated, ridges on inner sides of external and internal cotylae more developed; Dedicated to Dr. George M. Sutton on the occasion of his 70th birthday. 281 282 THE WILSON BULLETIN September 19611 Vol. 80, No. $ Fc. l. Stereophotographs of the type of Palaeophasianus meleagroides. Upper left, proximal end of tarsometatarsus; upper right, distal end of tibiotarsus; lower left, anterior view of tarsometatarsus; lower right, internal view of tibiotarsus. (6) slope of anterior margins from the top of the intercotylar prominence to the external and internal cotylae more gradual (from an anterior view); (7) shaft decidedly more triangular in shape (possibly due, in part, to crushing), the sides of the hypotarsns and shaft being more planar; (8) external cotyla somewhat less open anteriorly and pos- teriorly; (9) internal and external cotylae more round and cup-shaped; and (10) hypotarsus more developed, projecting more posteriorly. Tibiotarsus.--The fossil is similar to living 4ramus guarauna, but (1) from anterior view the internal condyle more elevated relative to external condyle (may be partially the result of crushing and displacement of bone); (2) internal condyle thicker basally, more triangular in shape (when viewed from distal end); and (3) rim of external condyle elevated more posteriorly relative to posterior rim of internal condyle (when viewed from distal end). Measurements. Tarsometatarsus: greatest breadth of head 18.5 mm; greatest depth of head (measured from tip of intercotylar prominence to most posterior portion of hypotarsus) 19.0 ram; width of shaft 30 mm below top of intercotylar prominence 12.3 joei EOCENE FOSSIL LIMPKIN 283 Cracraft mm; tibiotarsus: greatest breadth across condyles 16.5 mm; width of shaft 30 mm from top of internal condyle 9.6 mm; depth of shaft 30 mm from top of internal condyle 7.5 mm; greatest width of external condyle (measured from anterior to posterior) 16.0 ram. DISCUSSION The hypotarsus of the tarsometatarsus is badly damaged, but portions of several canals are still present. A well marked canal is found on the external side of the. hypotarsus, but it is impossible to say whether or not the canal was open or closed posteriorly. In addition, a larger, medial canal and a smaller, internal canal are present, but again, one cannot be sure whether they were grooves (i.e., open) rather than canals (i.e., closed). Taken by itself a positive identification of the fossil tibiotarsus is difficult. The bone was considerably damaged in preservation and portions of it were probably displaced as fossilization was taking place. Consequently, the above description, especially of the topographical relationships of the condyles to each other, may possibly be somewhat misleading. The fossil tibiotarsus superficially resembles that of tetraonids in some respects, for instance in the more developed, more triangular internal condyle. Unfortunately, the area of the supratendinal bridge is still covered by a very hard matrix and further preparation does not appear possible. The fossil does, however, resemble the Aramidae in general features, and there is no good reason for doubting its inclusion along with the tarsometatarsus in this family. The distal end of the tarsometatarsus included in AMNH No. 5128 still remains imbedded in matrix on one side, and the shaft and trochleae are so broken up, that if more matrix were removed, the fossil would break apart. Because of this situation, the distal end of the tarsometatarsus cannot be identified in itself. It is also not possible to identify the remaining fragments of the long bones. Three fossil aramids have been described from the early Tertiary: Badistornis aramus Wetmore White River series, Upper Oligocene, South Dakota Gnotornis aramiellus Wetmore White River series, Upper Oligocene, South Dakota 4ramornis longurio Wetmore Snake Creek Beds, Middle Miocene, Nebraska Palaeophasianus appears to resemble Badistornis in certain features, but when compared with Wetmore's description (1940), Palaeophasianus differs in the following characters: (1) internal cotyla is not as high relative to the external cotyla; (2) internal cotyla is more round; (3) external cotyla is apparently not as open anteriorly or posteriorly; and (4) comparison with 284 THE WILSON BULLETIN Scptember1968 Vol. 80, No. 3 Wetmore's Figure 7 indicates the anterior margin from the intercotylar prominence to the external cotyla is much less vertical (from an anterior view). All of these characters, along with geologic age differences, suggest that Palaeophasianus is generically distinct from Badistornis. Gnotornis is represented by the distal end of a left humerus. The measure- ments given by Wetmore (1942) indicate Gnotornis was approximately one- third the size of either fossil or living limpkins. On the basis of certain characters of the humerus, Wetmore considered Gnoto.rnis to be a distinct genus. Because only the humerus of Gnotornis is preserved, a comparison with Palaeophasianus cannot be made. zlrarnornis is represented by the distal end of a left tarsometatarsus. Accord- ing to the measurements (Wetmore, 1926) zlrarnornis was slightly larger than the Recent genus zlrarnus. A comparison of size between the type of zlramornis (AMNH No. 6292) and the damaged tarsometatarsus of Palaeo- phasianus shows that the latter is considerably larger. The tarsometatarsus of Palaeophasianus is badly damaged, hence a comparison with zlramornis cannot be made. Shufeldt (1915) placed another fossil (distal end of right tarsometatarsus) from the Bridget Formation of the middle Eocene of Wyoming in the genus Palaeophasianus. The fossil (Yale Peabody Museum No. 896) was compared to the types of Palaeophasianus and zlramornis and to skeletons of zlramus. This second specimen is so badly damaged--the trochlea for digit 2 is gone, the posterior side of the trochlea for digit 3 is lacking, and the trochlea for digit 4 is slightly broken--that comparison is difficult. However, there is little doubt that the Yale specimen is larger than Palaeophasianus melea- grogdes. Moreover, certain characters suggest this bone is not a limpkin: the distal foramen is farther removed proximally from the base of the trochlea for digit 3 and the external intertrochlear notch than in zlramus, and the base of the trochlea for digit 2 appears not to be directed posteriorly as it is in zlramus. The Yale specimen may possibly be an aramid, but because the bone is greatly damaged, positive identification is nearly impossible. Wetmore (1940:33) believed the differences of Badistornis from the Recent genus zlrarnus "tend to ally it to the cranes, the Gruidae, so that it appears ancestral to the modern limpkins. As it gives a closer approach to the cranes than does living Aramus it indicates more certainly the pre- supposed line of ancient connection between the Aramidae and the Gruidae." Palaeophasianus also resembles the Gruidae in some characters but no more so than it does several other families. The resemblances seem better explained on the basis of characters inherent in the tarsometatarsus and tibiotarsus themselves and appear not to be a reflection of relationship. Joel EOCENE FOSSIL LIMPKIN 285 Cracraft Included with the Aramidae and Gruidae in the superfamily Gruoidea is the Eocene family Geranoididae (Wetmore, 1933). The type species, Geranoides jepseni, is based on the fragmentary remains of the distal ends of a tarsometatarsus and tibiotarsus. The tarsometatarsus is distinctly different from that of the Aramidae. The tibiotarsus of P. meleagroides shows some differences from the tibiotarsus of Geranoides, notably in the shape of the external condyle. Due to the fragmentary nature of the type material of Palaeophasianus, comments about its relationship with Ge'ranoides are prob- ably best kept at a minimum at this time. The placing of Palae'op.hasianus in the Aramidae extends the known occurrence of that family back to the early Eocene and indicates that the family had attained a remarkable diversity by the early Tertiary. ACKNOWLEDGMENTS I am especially grateful to Dr. Malcolm C. McKenna for all the assistance he extended to me and for his critical reading of the manuscript. Dr. Alexander Wetmore gave freely of his time to read the manuscript and for this I am grateful. Dr. John Ostrum of the Peabody Museum of Yale University allowed me to borrow material in his. care. I want to thank also the authorities of the Department of Ornithology of the American Museum of Natural History for allowing me the use of their collections. SIIUFELDT, 1913 1915 WETMORE, 1926 1933 1940 1942 LITERATURE CITED BROOKORB, P. 1964 Catalogue of fossil birds. Part 2 (Anseriformes through Galliformes). Bull. Florida State Mus., 8:195-335. R.W. Further studies of fossil hirds with descriptions of new and extinct species. Bull. ,4met. Mus. Nat. Hist., 32:285-306. Fossil birds in the Marsh Collection of Yale University. Trans. Connecticut /Icad. /lrts Sci., 19:1-110. A. Descriptions of additional fossil birds from the Miocene of Nebraska. ,4mer. Mus. Novlt., No. 211. Fossil bird remains from the Eocene of Wyoming. Condor, 35:115-118. Fossil bird remains from Tertiary deposits in the United States. J. Morphol., 66: 25-37. Two new fossil birds from the Oligocene of South Dakota. Smithsonian Misc. Cull., Vol. 101, No. 14. DEPARTMENT OF BIOLOGICAL SCIENCES COLUMBIA UNIVERSITY NEW YORK NEW YOR 10027, 14 FEBRUARY 1967.