576 ShortCommunications [Auk,Vol. 99 the tongue by the bill removesthe nectar from the on the manuscript,C. E. Bayer,D. Nelson, and L. tongue; it is still not clear, however, how the nectar Shemshedinofskifor helpingduring data reduction, moves from the anterior portion of the bill into the and B. Thomasand R. Munsonfor providinguse of esophagus.The nectarmay adhereto the baseof the facilitiesat TuckerWildlife Sanctuary.Research was extruded tongue, being brought to the base of the supportedby grantsfrom the NationalScience Foun- bill when the tongue is retracted.Suction may help dation (DEB 77-15521),the Chapman fund of the transport nectar into the esophagus:our film shows AmericanMuseum of Natural History, the Orange a bulge in the throat region as the tongueis being County chapter of the Audubon Society, and the extended. Harry Frank GuggenheimFoundation. Among hummingbirds, long bills are believed to be beneficialbecause they facilitateprobing of flow- LITERATURE CITED ers with long corollas(Wolf et al. 1972). Our results provide a mechanism for such a benefit: at flowers EWALD,P. W., • S. ROHWER.1980. Age, coloration with long corollas,long bills may yield greaterrates and dominancein nonbreedinghummingbirds: a test of the asymmetryhypothesis. Behav. Ecol. of licking than short bills, becauselong bills permit Sociobiol. 7: 273-279. maintenance of small distances between nectar and HAINSWORTH,F. R. 1973. On the tongue of a hum- the bill tips, which squeezeloads of nectaroff of the mingbird: its role in the rate and energeticsof tongue.It is thereforenot surprisingthat the bills of feeding. Comp. Blochem.Physiol. 46: 65-78. hummingbirdstend to be similar in length to the , & L. L. WOLF. 1979. Feeding: an ecological corollasthat arevisited by the birds (Wolf et al. 1972, approach. Adv. Studies Behav. 9: 53-96. 1976). Such matching should result in short transit SCHARNKE,H. 1931. Beitr•ge zur Morphologie und times by the tongue between the nectar sourceand the inside of the bill. Entwicklungsgeschichteder Zunge der Trochil- idae, Meliphagidae und Picidae. J. Ornithol. 79: Although previouslymeasured negative correla- 425-491. tions between extractionrate and corolla length are WEYMOUTH,R. D., R. C. LASIEWSKI,& A. J. BERGER. most pronouncedwhen feederscontain unnaturally 1964. The tongue apparatusin hummingbirds. large volumes of food, such correlations are still sta- Acta Anat. 58: 252-270. tistically significantwhen food volumes are similar WOLF, L. L., F. R. HAINSWORTH, & F. G. STILES. to thosefound naturallyin flowers(Hainsworth and 1972. Energeticsof foraging:rate and efficiency Wolf 1979). This finding, coupled with our results, of nectarextraction by hummingbirds.Science suggeststhat long bills evolved, at least in part, be- 176: 1351-1352. causeincreased bill length increasesrates of licking , F. G. STILES, & F. R. HAINSWORTH. 1976. from flowers with long corollas.Long corollasprob- Ecologicalorganization of a tropical highland ably coevolvedbecause of pollination benefits asso- hummingbird community.J. Anim. Ecol. 45: ciated with specialization on pollinators (Wolf et al. 349-379. 1976). We thank C. E. Bayer,F. R. Hainsworth,B. Hallet, Received28 September1981, accepted 25 January1982. A. Heyneman, and R. D. Montgomeriefor comments

Tarsometatarsusof Protostrixfrom the mid-Eoceneof Wyoming

PAT V. RICH EarthSciences Department, Monash University, Clayton, Victoria 3168 Australia, and The NationalMuseum of Victoria,Melbourne, Victoria 3000 Australia Two, probably three (Rich and Bohaska 1976, frequently represented by this durable, hind-limb 1981), families of are known from Paleogene element. It is of some taxonomic interest, then, that sedimentsof North America and Europe, most taxa a distal fragment of a tarsometatarsusin the Verte- being representedby only one or two fragmentary brate Paleontologycollections of the American Mu- bones. One family, the Protostrigidae,which is en- seum of Natural History appearsto representPro- demic to North America, containstwo genera, Eos- tostrix. trix and Protostrix.Although severalspecimens (and In 1913 Shufeldt described the distal end of the species) of Protostrix are known (Brodkorb 1971), right tarsometatarsus(AMNH No. 2629, see Fig. 1) thus far no tarsometatarsusof any species in this among a number of bones that he proposed were genushas been recognized,and it has been difficult from "some medium sized falconine species" (p. to make comparisonswith other fossil owls, most 295), which had been collected in 1903 from the lower JULY1982] ShortCommunications 577

A B

C

lcm

Fig. 1. Protostrixcf. leposteus(AMNH 2629)from the mid-EoceneBridger Formation of Grizzly Buttes, Wyoming.A, posterior;B, anterior;and C, distalviews. Photographsby C. Tarka.

part of the mid-Eocene,Bridger Formation of the and outer trochleaethat extend far posteriorly, re- Grizzly Buttes,Green River Basin,Wyoming. He suitingin the highly archedcondition of the troch- alsomentioned the "upper part of the shaftof the leae in distal view. tarsometatarsus"and the "upper extremityof left Of owls known from the Paleogene,only seven coracold" that were "all from the same individual." speciesand four generaare representedby tarso- How he made this associationis not explained, and metatarsi and thus are directly comparableto the at the time of this studyI wasunable to locateeither GrizzlyButtes specimen: Ogygoptynx wetmorei (early of theseadditional two specimens.Thus, this paper Paleocene, North America), Eostrix miraica (early dealsonly with the distalright tarsometatarsus that Eocene, N.A.), E. raartinellii (Eocene, N.A.), Necro- he reported.Reexamination of this specimenhas byas harpax, N. rossignoli,N. edwardsi,and Asio shown it to be a strigiformthat sharesmore char- (originallyplaced in the genusOtus) henrici(all acterswith the EoceneEostrix (Wetmore 1938, Martin Eocene-Oligocene,Europe). and Black 1971) than with any other strigiform;it AMNH 2629 differs from the North American alsoexhibits unique characters of its own. forms as follows: Ogygoptynx(Rich and Bohaska AMNH 2629is clearlya strigiformas is indicated 1976, 1981) has an inner trochlea (II) that extends by the shallow,narrow middle trochlea (III) relative fartherdistally than trochleaIII, insteadof III being to the inner and outer trochleae;the extremelybroad longest;a trochleaIV thatis locatedfarther proximad inner trochleawith an elongatewing; the distal ex- on the shaft, terminatingat about the point where tensionof the inner trochleabeyond the outertroch- the proximalend of trochleaIII originatesinstead of lea (IV), beingexceeded only slightlyby trochleaIII; being distal to this origin; and in distal view, a the anteroposteriorcompression of the shaft with trochleaIV that is shiftedfarther posteriorlyrelative deep posteriorchannelling; and the elongateinner to other trochleae and an articular surface that is 578 Short Communications [Auk, Vol. 99

TABLE1. Comparisonof hind limb measurementsof early Tertiary owls of North Americaand Europe(in mm). Abbreviationsinclude: AMNH, Amer. Mus. Nat. Hist., Dept. Vert. Paleo., New York; KU, Univ. Kansas, Lawrence; MHNL, Mus. Hist. Nat., Lyon; MM, Munich Mus.; NMV, Natl. Mus. Victoria, Mel- bourne; PM, Mus. Natl. Hist. Nat., Paris; and YPM, Yale PeabodyMus., New Haven.

Eostrix Strigogyps mimica dubius Protostrix Protostrix USNM Eostrix MHNL AMNH leptosteus lydekkeri 15156, martinellii PQ 1073 Measurement 2629 YPM 512 AMNH 5165 14774 KU 16601 (cast)

Tibiotarsus 1. Transverse breadth acrosscondyles -- 13.5a -16.5 b 11.5 -- 21.1 2. Transverse breadth at distal end of shaft -- -10.9 b 11.9 • 8.9 -- -- 3. Maximum antero~posterior diameter of condyles -- >9.8 ½ 12.9• 9.9 -- 30.3 tarsometatarsus 4. Maximum width of distal end 14.3 -- -- 13.2 9.8 -- 5. Width of trochlea III •4.3 -- -- 54.8 ½ 3.5 -- 6. Depth of trochleaI! 6.3 -- -- •5.2 c 6.1 -- 7. Depth of trochleaIII 5.8 -- -- 6.1½ 4.3 -- 8. Depth of trochlea IV 8.8 -- -- •6.5 c -- --

From Wetmore (1937:85). Measuredfrom illustrationor givenin Wetmore(1933). Measured from castof specimen.

broaderand flatteneddistally. Eostrix(based mainly er than the internal border and the external border on E. martinellii, KU 16601) (Wetmore 1938, Brodkorb is not straight but decidedlyconvex externally;a 1971, Martin and Black 1972) has an asymmetric trochleaIV that is decidedlydeeper and more re- trochleaIII with the lateralborder decidedlydeeper curved internally; trochleaethat are more highly than the internal border, unlike the symmetrical arched; and a distal end that is deeper and more trochlea in AMNH 2629, and a trochlea IV that ap- mediolaterallycompressed. pears to be narrower and more anteroposteriorly Clearly, AMNH 2629 exhibits more overall simi- elongatedand deeper. larity to the North AmericanProtostrigidae than to Of the EuropeanPaleogene owls, all speciesof any othergroup, but it is distinctfrom Eostrix. It was Necrobyas(N. edwardsi,N. harpax,N. rossignoli)as recoveredfrom the same area in the Grizzly Buttes well as Asio henrici are smaller than AMNH 2629 near Fort Bridger,Wyoming, as was the type tibi- (Milne-Edwards1892; Galllard 1908, 1938). Strigo- otarsus (YPM 512) of Protostrix leptosteus(Marsh gyps dubius (Gaillard 1908) is decidedly larger. 1871, Wetmore 1937), and is within the size range AMNH 2629further differs as follows:N. harpax,N. expectedfor this large protostrigid(see Table 1), rossignoli,and N. edwardsihave a relatively broader beingsmaller than expected for P. lydekkeri(Shufeldt trochlea III; trochlea IV is more recurred, not 1913),if oneassumes that proportions of Eostrixmim- straight,along the lateralmargin (when viewed dis- ica are reasonableones to expectfor Protostrix.Of tally); trochleaeare more highly arched;and the dis- course,this may be an invalid assumption;until an tal end is decidedly deeper, relative to width, pro- associatedspecimen of Protostrixis found, this as- ducinga deep, more enclosedtendinal canal. In both sumption cannotbe demonstratedconclusively. It N. harpaxand N. edwardsithe posteriorpart of troch- seemsreasonable, then, that AMNH 2629 probably lea II, when viewed distally, is decidedly narrower representsthe remainsof P. leptosteus,which rein- than the anterior,unlike the broad posteriorwing in forcesthe suggestionthat Protostrixand Eostrix,al- Protostrix; in anterior view, the proximal end of thoughsimilar in manycharacters, are distinct with- trochleaI! lacks the prominent internal projection in the Protostrigidae,and that the protostrigidsare presentin Protostrix;trochlea II projectsabout as far distinct from both the North American Ogygoptynx distad as trochleaIII, whereasIII is distinctlylonger and the Europeanowls of Paleogeneage. in AMNH 2629. Thanks are due Drs. M. McKenna (Amer. Mus. Asio henrici has a more robust trochlea II with a Nat. Hist., New York), L. Martin (Univ. Kansas, narrowerwing (vieweddistally); a moreasymmetric Lawrence),J. Ostrom(Yale Univ., New Haven), D. trochleaIII, where the externalborder is much deep- Russell (Mus. Natl. His. Nat., Paris), and Mr. A. JuLY1982] ShortCommunications 579

TABLE 1. Continued.

Necrobyas Necrobyas Asio henrici Ninox Necrobyas harpax rossignoli Bubo novaesee edwardsii Milne- Milne- Milne- virginianus landiae Gaillard Edwards Edwards PM Qu PM Qu PM Qu Edwards NMV NMV 1938 1892 1892 3117 3118 3119 1892 B9788 B12401

•.2 15.7 8.8

6.7 12.1 7.0

7.1 12.1 7.6

11.0 10.0 9.3 9.1 8.8 9.0 16.9 10.5 ------3.8 3.8 -- 6.5 4.1 ------5.1 5.3 -- 8.6 5.6 ------4.0 4.1 -- 8.0 4.1 .... 6.4 -- 10.3 6.4

McEvey (Natl. Mus. Victoria, Melbourne) for the sud de la FranceC.R. DeuxiemeCongr'es Inter- loan of material. S. Olson, T. Rich, and A. McEvey nat. Ornithol. Budapest:60-80. criticized and M. L. Vickers and E. S. Pullum typed RICH, P. V., & D. J. BOHASKA.1976. The world's the manuscript.Chester Tarka beautifully photo- oldestowl: a new strigiformfrom the Paleocene graphed and prepared Fig. 1. of southwestern Colorado. Smithsonian Con- trib. Paleobiol. 27: 87-93. LITERATURE CITED --, & . 1981. The Ogygoptyngidae,a new family of birds from the Paleoceneof North Brodkorb,P. 1971. Catalogueof fossilbirds. Part 4 America. Alcheringa3: 95-102. (Columbiformesthrough Piciformes). Bull. Flor- SHVFELDT, R. W. 1913. Further studies of fossil ida Mus. 15: 163-266. birds with descriptions of new and extinct GAILLARD,M. C. 1908. Les oiseaux des Phosphorites species.Bull. Amer. Mus. Nat. Hist. 32: 285306. du Quercy. Ann. Univ. Lyon, N.S. I, Sci., Med- WETMORE,A. 1933. The statusof antiqua, icine 23: 1-178. Aquilaferox and lydekkerias fossil birds. . 1938. Contribution a l'•tude des oiseaux fos- Amer. Mus. Novitates 680: 1-4. siles. Arch. Mus. Hist. Nat. Lyon 15: 1-100. --. 1937. The systematicposition of Bubolep- MARSH, O. C. 1871. Notice of some new fossil mam- tosteus Marsh. Condor 39: 84-85. mals and birds from the Tertiary Formation of --. 1938. Another fossil from the Eocene of the West. Amer. J. Sci. Ser. 3, 2: 120-127. Wyoming. Proc. U.S. Natl. Mus. 85: 27-29. MARTIN, L. D. & C. C. BLACK. 1972. A new owl from the Eoceneof Wyoming. Auk 89: 887-888. Received7 May 1980, resubmitted10 September1981, MILNE-EDwARDS, g. 1892. Sur les oiseaux fossiles accepted11 November1981. de-d•p6ts •oc•nes de Phosphate de Chaux du