FOSSIL FLORA. 721 



This fine species belongs certainly to the genus Magnolia, as attested 

 by the shape and the forking, camptodrome secondaries. It is, however, 

 quite unlike any of the other species found in the Yellowstone National 

 Park. Perhaps its closest relation is M. californica Lx., 1 from the Chalk 

 Bluffs of California. It is different in shape, being ovate instead of broadly 

 oval, and has somewhat different secondaries. The large quadrangular 

 finer nervation is similar in both. 



It does not approach very closely to either of the living American 

 species, being perhaps closest to M. acuminata L., the well-known cucumber 

 tree. 2 The shape of the leaves is practically the same, but the nervation 

 differs somewhat. 



I have named this species in honor of Prof. Ceorge E. Culver, who 

 assisted in making the collection from this place. 



Habitat: East bank of Lamar River, between Cache and Calfee creeks; 

 collected by F. H. Knowlton and George E. Culver, August, 1888. 



Magnolia ? pollardi n. sp. 

 PI. LXXXI, figs. 9, 10. 



Petals of firm texture, elliptical or elliptic-ovate in outline, narrowed 

 below, rounded-obtuse above; nervation of numerous approximately par- 

 allel nerves about 2 mm. apart. 



The best preserved of these 2 specimens (fig. 9) has 7.5 cm. in length 

 preserved, and was probably fully 8.5 cm. in length when perfect. It is 

 3 cm. broad in the middle, and is narrowed at base to a point of attach- 

 ment some 5 or 6 mm. broad. The upper point is unfortunately destroyed, 

 but it seems probable, from the appearance of the margin and nerves, that 

 it was obtuse. The nerves arise from the basal part and run approximately 

 parallel, spreading slightly in the middle and converging toward the apex. 

 In the middle these nerves are between 3 and 4 mm. apart, but in the apex 

 they are separated by only about 2 mm. There is some evidence of inter- 

 mediate nerves, or possibly cross veinlets, but these are so indistinct that 

 a positive statement concerning - them can not be made. 



The other specimen (fig. 10) is a trifle over 5 cm. in length, but lacks 

 both upper and lower parts. It was probably 6.5 or 7 cm. in length when 



'Mem. Mus. Comp. Zocil., Vol. VI, No. 1, p. 25, PI. VI, figs. 5, 7. 

 J Cf. Sargent: Silva of N. A., Vol. I, Pis. IV, V. 

 MON XXXII, PT II 40 



