128 BULLETIN 110, UNITED STATES NATIONAL MUSEUM. 



In 1903 Dr. H. F. Osborn' described and figured a well-preserved skeleton from 

 the famous "Bone Cabin" quarry in Wyomimg, as the type of the new genus 

 and species Ornitholestes hermanni. This specimen comes from approximately the 

 same geological level as the type of Coelurus fragilis Marsh, and only a few miles 

 distant from where the type was found. Osborn distinguished OrnitJiolcstes from 

 Coelurus by the "nonserration of the teeth, by the relatively short cervical verte- 

 brae, by the less extreme hollowness of all of the vertebrae." 



A review of the material on which Coelurus was based renders the distinction 

 made less obvious. Although the genus Coelurus was established in 1879 no men- 

 tion was made of the presence of teeth until 1896,^ when M^rsh figured a tooth as- 

 cribed to Coelurus fragilis (see fig. 4, pi. 34). Lull, in a letter of June 28, 1915, has 

 been kind enough to furnish me with the following information concerning this 

 tooth : 



The tooth, figure 1 [fig. 4, pi. 34], was entirely disassociated from the rest of the material and bears 

 the acces.sion number | i271 |, being collected by Reed in 1879 from Como. Wyoming; but 1 can not tell 

 from which quarrj'; but, as you will see, the tooth came in some time in advance of the rest of the ma- 

 terial [refers here to the figured specimens]. Hence there is no indication of the as.sociation. As the 

 original description Is based on the vertebrae, the association of the serrated tooth is highly conjectural. 



In this connection it is of interest to quote further from Lull's letter regarding 

 the bones figured by Marsh (on pi. 7) in his Dinosaurs of North America : ' ' Ver- 

 tebra, figure 2, is catalogue number 1993. Caudal, figure 4, is catalogue number 

 1992." Osborn's distinction "by the relatively short cervical vertebrae " in Orni- 

 tholestes can hardly be considered a good distinctive character unless found to exist 

 in an articulated series. And in Ornitholestes only two cervicals were ]>Tesent, and 

 as has been shown above the cervical figured by Marsh does nor pertain to the type 

 specimen. Furthermore, it is now known that the relative length of the cervicals 

 vary with the position they occupy in the cervical series. Articulated cervical 

 series in Antrodemus and Ceratosaurus show the posterior members to be more 

 elongated than the anterior cervicals and this is probably true also of the forms 

 under discussion. 



The extreme hollowness of the vertebrae is a difference of degree and certainly 

 not a character u})on which to base generic distinctions. Since a comparison of 

 the vertebrae of the type specimen Coelurus with those of the skeleton on which 

 Ornitholestes is based show only minor differences I do not see that their generic 

 distinctness has been established. 



COELURUS AGILIS Marsh. 



Plate 34, figs. 5 and 6. 



Coelurus agilis Marsh, Amer. Journ. Sci,, vol, 27, 1884, p, 335, pi. 10, figs. 3a, 35,- Sixteenth Ann. 

 Kept. U. S. Geol, Survey, 1896, pt, 1, pi. 10, figs. 3, 4,— Zittel, K. A,, Handbuch der Palaeon- 

 tologie, vol, 3, 1890, p, 732.— Hay, Bull. No. 179, U. S. Geol. Survey, 1901, p. 493.— Nopcsa, 

 F. B., Foldtani Kozlony, Budapest, vol. 31, 1901, p, 202.— Gilmore, Bull. 89, U. S, Nat, Mus,, 

 1914, p, 25,— MooK, C, C, .\nn. New York Acad. Sci., vol, 27, 1916, p, 142, 



A pair of articulated pubes (No. 5737, U.S.N.M.) from which portions of the 

 proximal articular ends are missing is here provisionally referred to the present 



1 Bull. Amer. Mus. Nat. Hist., vol. 19, 1903, pp. 459-464, figs. 1, 2, 3. 

 2 16th Aim. Rcpt. U. S. Geol. Surv., pt. 1, 1S96, pi. 7, flg. 1, 



