1893.] DR. C. J. FORSYTH MAJOR OK MIOCENE SQUIRRELS. 211 



examination o£ the figures and descriptions without having seen 

 the originals ; but I wish to offer a few remarks. 



Marsh has considered a certain number of isoLated molars, 

 possessing three longitudinal pairs of elevations, to be upper molars, 

 although the type of Dipriodon robustus, the only molar which is 

 undoubtedly from the upper jaw, left side, " its position being 

 decided by a portion of the maxillary attached to it," ^ appa- 

 rently has only two longitudinal rou's. This circumstance has 

 given origin to part of Osborn's criticisms'^. The type of Tripri- 

 odon ccelatus^, considered by Marsh ■* as the first upper molar of the 

 left side, as well as the type of Selenacodon fragilis '% also stated to 

 be an upper molar (both having three longitudinal rows of cusps), 

 are declared by Osborn to be respectively a last lower molar of 

 Meniscoessus, and an anterior lower molar of the same ^. In a 

 subsequent note ^ Osborn writes as follows : — " It remains for the 

 author to show specifically that the types of Selenacodon and 

 Tripriodon are maxillary teeth," adding : " I should myself have 

 considered them as such but for the fact that the type of Dipriodon 

 robustus, with two rows of tubercles, was described as a maxillary 

 tooth, and figured with a supposed fragment of the zygomatic 

 arch attached to the alveolar border." With regard to the upper 

 molars, Marsh asserts, in his latest paper on the subject, that he 

 has the means of showing what Osborn has objected to : " Although 

 not found in position in any one specimen, so many have been 

 secured with portions of the jaw attached, that their place in the 

 dental series has been ascertained in several forms ; " and he goes 

 on to state, " that the upper molar teeth may be separated into two 

 series, the first having three longitudinal row s of elevations on the 

 crown, and the second series but two ro\A"s " ^ AVith the caution 

 imposed by the fact that I am judging only from the published 

 figures, I venture to suggest that the type of Biprlodon robustus, 

 which has the undoubted fragment of the zygomatic arch attached 

 to the alveolar border, had originally three longitudinal rows of 

 cusps, the middle one being worn off. Marsh himself states that 

 its " points ai'e somewhat worn"', and this appears to me clearly 

 shown in his figure '", If we now assume that x^here there are 

 three rows of tubercles above and two below, " the cusps of the 

 lower rows fit into the valleys of the upper teeth'' " (which in my 



1 Marsb, I. c. part I. p. 85, pi. ii. figs. 1.3-15. 



^ As stated by O.sborn binisell' in " A Reply to Professor 0. C. Marsb'.s ' Note 

 on Mesozoic Mammalia.' " Reprinted with sligbt alterations from the 'American 

 Naturalist; September 1891, p. 782. 



3 Marsh, I.e. part I. pi. ii. figs. 19-21. 



' L. c. p. 80. 



■' Marsh, /. c. part I. pi. ii. figs. 22-24, p. 86. 



" H. Fairfield Osborn, "A Review of the Cretaceous Mammalia" (Proc. Acad. 

 Nat. So. Philadelphia, 1891, p. 128). 



■^ A Reply to Pr(jCe.ssor O. C. Marsh's Note, &c., p. 782. 



" Part III. ]). 253. 



« Part T. p. 85. 

 '" L.c. pi. ii. figs. 1.3, 14. 

 '' Osborn, " A Review of the Cretacoou.s Manimalia." 



14* 



