THE genus Telmatornis was described for two species, T. priscus and T. affinis, from the Late Cretaceous (Maestrichtian) of New Jersey (Marsh, 1870). In his original description Marsh concluded that Telmatornis was probably related to rails (Rallidae). Some years later Shufeldt (1915) described a third species, T. rex, which he also considered to be allied to the rails. Both Marsh and Shufeldt noted certain re- semblances between Telmatornis and the herons and larger shorebirds, but Shufeldt was definite in his opinion that Telmatornis was not a heron. Subsequent authorities (Lambrecht, 1933: 489; Wetmore, 1956: 62; Brodkorb, 1967: 116) have continued to place Telmatornis tenta- tively in the Rallidae. While undertaking a revision of the extinct genera of rails it was necessary for me to reexamine the systematic position of Telmatornis. After comparison with numerous genera of Recent and fossil rails and with nearly all families of nonpasserine birds, I have come to the con- clusion that Telmatornis is not referable to the Rallidae but is instead more closely related to the charadriiform family Burhinidae. In this paper I discuss this relationship and comment on the systematics of fossil burhinids. COMPARISON WITH THE RALLIDAE The humeri of Telmatornis (Figures 1 and 2) differ from those of rails in that: (1) the entepicondyle projects much less distally and anconally, (2) in palmar view, the external condyle extends less proxi- mally, less elongated proximodistally, (3) the attachment of M. pronator brevis is more distinct and developed into a deeper pit which is located much more palmarly, (4) the internal condyle is less round on its palmar surface (due to presence of a small depression), (5) the external tri- cipital groove is shallower, (6) the olecranal fossa is much shallower, (7) the ridge between the internal condyle and entepicondyle is not con- stricted or depressed, (8) the impression of the brachialis anticus is lo- cated less internally, and (9) in external view, the shaft is curved more distally and proximally and is more S-shaped. AFFINITIES WITH THE CHARADRIIFORMES AND BURHINIDAE The preceding comparison demonstrates a distinct morphological dif- ference between Telmatornis and rails. On the other hand, Telmatornis agrees much more closely with the family Burhinidae of the Charad- 36 The Auk, 89: 36-46. January 1972 Figure 1. Telmatornis priscts, type, 3/4PM 840. Lower, stereophotographs of palmar view. Upper, stereophotographs of distal end. Both about X 1.5. riiformes. It is difficult to characterize the Charadriiformes solely on the basis of the distal end of the humerus, but the order can be divided into three morphological groups: (a) the Alcidae, (b) the Burhinidae, and (c) all other families. The Jacanidae are somewhat intermediate between groups (b) and (c). In general the distal end of the charadrii- form humerus tends to have the following characters: (1) the attach- ment of M. pronator brevis is formed into a pit (sometimes deep) that is located far palmarly, (2) the ectepicondylar prominence is well-de- veloped (usually into a long process), (3) the entepicondyle typically does not project much distally relative to the internal condyle (but can project anconally), (4) the impression of M. brachialis anticus is moderately deep (and well-defined) within the brachial depression, which is extensive distally, (5) typically the shaft narrows sharply away from the distal end so that the distal end appears broad relative to the width of the shaft, (6) the internal tricipital groove is well-developed, and (7) the olecranal fossa tends to, be deep and formed into a slight shelf set off from the shaft rather than grading smoothly into the shaft. Except as noted below, Telmatornis agrees with this generalized description of the charadriiform humerus. Telmatornis can be grouped with the Burhinidae on the basis of the following shared characters: (1) the area between the ectepicondyle and ectepicondylar prominence is the same shape, and the pits of the Figure 2. Telmatornis a]]inis (lower), type, YPM 845, and Telmatornis rex (upper), type, YPM 902. Stereophotographs of palmar view. Both about X loS. tendons and ligaments are in similar positions, (2) the attachment of M. pronator brevis is formed into a deep pit and is located palmarly, (3) in ventral view, the tendon pits on the entepicondyle have similar shapes and relative positions, (4) the entepicondyles do not project much distally or anconally, (5) the olecranal fossa is shallow, similar in shape, and grades rather smoothly into the shaft, and (6) the internal condyle is not bulbous but rather elongated lateromedially and projects distally only moderately relative to the external condyle. Although the above shared characters unite Telmatornis and the burhinids within the Charadriiformes, a detailed comparison indicates that Telmatornis is as distinct from the burhinids as are some other taxa of family rank (e.g. the Jacanidae). Hence, it is desirable to recognize a new family, to be called TELMATORNITHIDAE, new family TYPE GENus: Telmatornis Marsh. DIAGNOSIS: In characters of the humerus, the Telmatornithidae differ from the Burhinidae as follows: (1) the internal condyle is less rounded on the palmar surface because of the presence of a small depression, (2) the internal and external condyles are located closer to each other, (3) the external tricipital groove is shallower, (4) in external view the shaft is curved more at the distal end, (5) the ectepicondylar prominence protrudes less externally, and (6) the distal portion of the entepicondyle is broader internoexternally. Genus TELMA TORNIS Marsh Telmatornis Marsh, 1870, p. 210 TYPE SPECIES: Telmatornis priscus Marsh. DIAGNOSIS: Same as for family; only included genus. Telmatornis priscus Marsh Figure 1 Telmatornis priscus Marsh, 1870, p. 210 HOIOTYPE: YPM 840, distal end of right humerus; from upper Cretaceous deposits (middle Maestrichtian in age), Navesink Forma- tion ("middle marl bed"; Baird, 1967); Cream Ridge Marl Company quarry near Hornerstown, Monmouth County, New Jersey. MEASVE>aENTS: See Table 1. RElYtAft, S: See below under T. affinis. Telmatornis affinis Marsh Figure 2 Telmatornis affinis Marsh, 1870, p. 211 HOLOTYPE: YPM 845, distal end of right humerus; from upper Cretaceous deposits (middle Maestrichtian in age), Navesink Formation ("middle marl bed"; Baird, 1967); Cream Ridge Marl Company quarry near Hornerstown, Monmouth County, New Jersey. MEASUE>aENTS: See Table 1. REVtAKS: Telmatornis affinis is very similar to T. priscus in both size and morphology. Marsh (1870: 211) held that T. affinis was a separate species from T. priscus because the "notch between the radial and ulnar condyles is somewhat deeper; the elongated tubercle, on the inner surface behind the notch, which confines the upper tendon of the triceps muscle, is larger; the impression of the anterior brachial muscle JOEL CRACRAFT [Auk, Vol. 89 on the outer surface is higher up, and more shallow; and the epitrochlear elevation is more prominent." The only difference I note in the two species is the slightly larger size (and therefore greater robustness) of T. priscus (Table 1). The other characters mentioned by Marsh are either not apparent to me or are too slight to accept as species differences. Both bones appear to be those of adult birds. The only justification for separating these species, then, is size, and in this instance the differences are probably not sig- nificant. Hence, I recommend synonymizing T. affinis and T. priscus; the latter name would apply to the species because of page priority. Telmatornis rex Shufeldt Figure 2 Telmatornis rex Shufeldt, 1915, p. 27 HOIOTYPE: YPM 902, distal end of right humerus; from upper Cretaceous deposits (probably late Maestrichtian in age); Hornerstown marl; Monmouth County, New Jersey. RFRRr> M^TR^L: YPM 948, distal end of left humerus; from up- per Cretaceous deposits (late Maestrichtian in age); locality data on box gives "Cream Ridge Marl Co." quarry; Monmouth County, New Jersey. M^surNTs: See Table 1. RvrKs: The two specimens of T. rex show slight differences in size. The type is somewhat broader across the condyles and the referred specimen has a slightly heavier shaft. The brachial depression of the type is also less well-defined. If the same criteria used to separate T. priscus and T. afJinis were followed here, then the two specimens of T. rex could not be considered conspecific. However I favor uniting them under one name. Except for its decidedly larger size, T. rex does not differ from T. priscus in any significant features. DISCUSSION SYSTEMATIC POSITION O1 MILNE.4 LYDEKKER, 1891 Because Milnea gracilis (Figure 3), described for a humerus from the Aquitanian (late Oligocene-early Miocene) of France, is the only pre-Pleistocene fossil currently placed in the Burhinidae, its correct systematic position is central to a discussion of the relationships of Tel- matornis. A comparison of the type of Milnea gracilis (BM(NH)47457) with Burhinus and representatives of other nonpasseriform families shows that Milnea should be allocated to the Threskiornithidae rather Figure 3. Milnea palmar view. Upper end. All about X 1.2. gracilis, type, BM(NH) 47457. Lower, stereophotographs ot right, anconal view distal end. Upper, anconal view, proximal than to the Burhinidae. The following differences that Milnea exhibits from Burhinus will serve to demonstrate the significant morphological gap between these two taxa: (1) the humerus is decidedly heavier and more robust, (2) the external condyle is relatively smaller, (3) the olecranal fossa is less deep, (4) the area distal to the impression of M. brachialis anticus is less depressed and excavated, (5) the entepicondyle is less pronounced, (6) the pneumatic fossa is much smaller, (7) the capital groove and area distal to the head (on anconal surface) is much less excavated, (8) the external tuberosity is less pronounced, (9) the ridge extending distally down the shaft from the median crest is less pronounced, and (10) the bicipital surface is less inflated, less bulbous, and flatter. When Milnea is compared with the threskiornithids, numerous shared characters indicate a close relationship. Among these features are: (1) a deep, oblong-shaped impression of M. brachialis anticus, (2) condyles of the same shape and position, (3) same shape and contour of the shaft, especially the distal end where the tapering is distinctive, (4) rather pointed humeral head, (5) moderately deep and elongated capital groove oriented more or less perpendicularly to the long axis of the shaft and noticeably undercutting head, (6) thin median crest, (7) small pneumatic foramen, and (8) internal side of shaft just proximal to attachment of M. pronator brevis constricted anconal-palmarly. Three European pre-Pleistocene species are currently assigned to the Threskiornithidae (Brodkorb, 1963a: 277-278). Ibidopsis hordwelliensis from the late Eocene of England is incorrectly placed in the Threskior- nithidae and is instead a rallid (Cracraft, MS). A species from the Aqui- tanian of France, Ibidopodia palustris Milne-Edwards, is represented by a cranium and tarsometatarsus and thus cannot be compared directly with Milnea. Another threskiornithid species from the Aquitanian of France, Eudocimus paganus (Milne-Edwards), compares closely with Milnea gracilis. The specimens of E. paganus that I have examined (USNM 6360, 6361; includes four specimens) differ greatly in size and mor- phology, and the humerus of Milnea probably could be included within this variation (Table 2). The humeral head of Milnea is more pointed but this could reflect individual variation; the distal ends of Milnea and E. paganus are very similar. Two size ranges may be present within the specimens of E. paganus but only a statistical analysis of the abundant material in European collections will provide evidence about specific limits. The present data suggest Milnea gracilis is conspecific with Eudocimus paganus or at least very closely related to it. COMPARISON AND PHYLOGENETIC SIGNIFICANCE OF TELMATORNIS In addition to Telmatornis priscus and T. affinis Marsh (1870) also described three species of birds from upper Cretaceous deposits (Navesink Formation) of New Jersey and placed them in a new genus Palaeotringa. This genus is currently allocated to a separate subfamily within the Scolopacidae (Brodkorb, 1967). Two species, P. vagans and P. vetus, are very fragmentary and probably cannot be identified even to order. The third species, P. littoralis, is based upon a distal left tibiotarsus with the internal condyle lacking. Shufeldt (1915: 23) considered the tibio- tarsus to represent a gull and not a wader, but because of the frag- mentary nature of this fossil allocation to a family may prove impossible. In any case, there is evidence that P. littoralis was not close to the Burhinidae and, by inference, to the Telmatornithidae. For example, 44 Jo Cu^cu^ [Auk, Vol. 89 January 1972] New Charadriiform Family 45 compared to Burhinus the external condyle of P. littoralis is not con- stricted as much proximally (in anterior view) and the short ridge which runs proximodistally from the supratendinal bridge is much more de- veloped. Hence, it is probable that Telmatornis and Palaeotringa do not have a close (i.e. familial) relationship. Any statement regarding the phylogenetic interpretation of Telmatornis must be considered tentative. The inclusion of Telmatornis within the Charadriiformes and near the Burhinidae suggests that the burhinids themselves, or at least their relatives, may be primitive within the order. This conclusion gains some support from a study by Brodkorb (1963b) that included several late Cretaceous charadriiforms from Wyoming. The family Cimolopterygidae is known for two genera, Cimolopteryx and Ceramorns, which Brodkorb believes are closest to the Recurvirostridae but which may also have burhinid or glareolid affinities. Telmatornis is definitely closer to the burhinids than to the recurvirostrids, but pos- sibly the Telmatornithidae and Cimolopterygidae represent an early radiation that eventually gave rise to the two modern families. Un- fortunately no humeri of the cimolopterygids are known, and thus a comparison with Telmatornis is impossible. The Telmatornithidae and Cimolopterygidae are both Maestrichtian in age and hence are the oldest members of the Charadriiformes. ACKNO,VLEDGMENTS I thank Alan J. Charig and Cyril A. Walker, British Museum (Natural History); Clayton Ray, United States National Museum; and John Ostrom and Peter Gaiton, Peabody Museum of Natural History, Yale University, for permission to study speci- mens in their care. I am grateful to Walter J. Bock and Lester L. Short for com- ments on the manuscript. This study was supported by a Frank M. Chapman Fellow- ship, 1969-1970, from the American Museum of Natural History. SUMMARY The Cretaceous genus Telmatornis is removed from the Rallidae and placed in the Charadriiformes near the Burhinidae. A new family, the Telmatornithidae, is erected for the genus. Telmatornis priscus and T. affinis are considered conspecific. The Aquitanian fossil Milnea gracilis is removed from the Burhinidae and is probably conspecific with Eudocimus paganus of the Threskiornithidae. LITERATURE CITED BAIrn), D. 1967. Age of the fossil birds from the greensands of New Jersey. Auk, 84: 260-262. BRODKORB, P. 1963a. Catalogue of fossil birds: Part 1 (Archaeopterygiformes through Ardeiformes). Bull. Florida State Mus., 7: 179-293. BRODKORB, P. 1963b. Birds from the upper Cretaceous of Wyoming. Proc. 13th Intern. Ornithol. Congr.: 55-70. BROmCORB, P. 1967. Catalogue of fossil birds: Part 3 (Ralliformes, Ichthyornithi- formes, Charadriiformes). Bull. Florida State Mus., 11: 99-220. L^rBRECT, P. 1933. Handbuch der Palaeornithologie. Berlin Gebrilder Born- traeger. LVDZCCZR, R. 1891. Catalogue of the fossil birds in the British Museum (Natural History). London, Brit. Mus. (Nat. Hist.). M^RSa, O. C. 1870. Notice on some fossil birds from the Cretaceous and Tertiary formations of the United States. Amer. J. Sci., 49, 2nd Ser. 205-217. SaUVLVT, R. W. 1915. Fossil birds in the Marsh Collection of Yale University. Trans. Connecticut Arts Sci., 19: 1-110. WZTrORE, A. 1956. A check-list of the fossil and prehistoric birds of North America and the West Indies. Smithsonian Misc. Coil., 131: 1-105. Department of Ornithology, the American Museum of Natural History, Central Park West at 79th Street, New York City, New York 10024. Present address: Department of Anatomy, University of Illinois at the Medical Center, Chicago, Illinois 60680. Accepted 19 August 1970.