FOSSIL FLORA. 719 



and some of the smaller ones 12 or 15 cm. long and 4 to 6 cm. wide. The 

 leaves are thick and leathery, and evidently belonged to an evergreen 

 species. 



It is altogether probable that the leaves obtained by Mr. W. H. Holmes 

 in 1878 from Amethyst Mountain and identiiied by Lesqnereux as Magnolia 

 lanceolata Lx.,' really belong to this species. As nearly as can be made 

 out from Holmes's description of the locality," it is the same as that which 

 aftbrded the specimens under discussion. But a careful comparison of these 

 numerous leaves with the figures given by Lesquereux, as well as with 

 specimens from the Auriferous gravels, makes it certain that they can not 

 belong to M. lanceolata. Magnolia spectahilis differs in being broader, more 

 rounded at base, with secondaries more curved and with numerous inter- 

 mediate secondaries. A still greater point of difference is in the texture of 

 the leaf Of M. lanceolata, Lesquereux says:^ "This leaf is not coriaceous, 

 rather of a thin suljstance," while 31. .spectahilis is thick and distinctly coria- 

 ceous or leathery. The finer nervation is not preserved in M. lanceolata, 

 so it is not possible to compare that point. 



From further evidence it appears that these identical specimens were 

 again submitted to Lesquereux in 1887, and he then identified them with 

 M. inglefiehli Heer,* a s])ecies that lie has also reported from Lassen County, 

 California, Green River group, etc. It is certainly much more closely 

 related to this than to M. lanceolata, as may be seen from Heer's figures^ and 

 specimens identified with it from California. It i.s of the same shape and 

 size as M. spectahilis and is described as being coriaceous, but it difters 

 somewhat in having the secondaries more scattered, apex irregular, etc. 

 The finer nervation also differs. They are undoubtedly close, but seem to 

 be sufficiently distinct for specific separation. 



Among living species the affinity of M. spectahilis is unquestionably 

 with M. gramliflom L., or M. fietida Sargent, as it is now called. The size, 

 outline, texture, and nervation are practically the same. 



According to Sargent,'' the direct ancestor of Magnolia fatida was 



I Mem. Mus. Comp. Zool., Vol. VI, p. 24, PI. VI, fig. 4. 



"-Twelfth Ann. Eept. U. S. Geol. and Geog. Surv. Terr , 1878 (1883), Pt. II, p. 49. 



^Loc. cit., p. 24. 



»Ct'. Proc. U. S. Nat. Mus., Vol. X, 1887, p. 46. 



6F1. Foss. Arct., Vol. VII, 1883, p. 121, PI. LXIX, fig. 1 ; PI. LXXXV, fig. 3; PI. LXXX^^1, fig. 9. 



<>Silva of North America, Vol. I, p. 3. 



