26 LEAVES OF LIRIODENDRON HOLM. 



rudimentary proiougatiou of the midrib, as I have mentioned above 

 and figured in the leaves of the living tree. It seems, then, as if we 

 have the same fact here, that the leaves show an original longer-leaved 

 type. But, owing to the shape of the notch in these three varieties, 

 whatever this is, more or less deeply and sharply sinuate, Heer did not 

 consider this circumstance as being of any essential importance, since 

 he united a form as Fig. 13a on Plate xlv {I. c.) of Marcouana with an- 

 other one, Fig. 3 on Plate xxiii {I. c), of the same variety, and he has 

 certainly considered these two as identical with the best reason. For 

 if we look at the small leaves. Figs. 6, 10, and 16 on Plate v, all 

 borne on living trees, we shall find the same manner of variation. But 

 I hardly consider these three forms as " varieties," in consequence of 

 the numerous gradation forms, which I have illustrated, and will again 

 call attention to some of these figures. In regard to the relative 

 proportions of length and breadth there is a series of gradations 

 from Figs. 35 and 36 to Figs. 16 and 41, and we can see the same in re- 

 gard to the depth of the notch at the apex of these leaves. I do not 

 think it would be too hazardous to suppose Heer's varieties as merely 

 forms, belonging to trees or branches of some different age, either very 

 young or perhaps more mature. But Dr. Newberry objects even here, 

 because Heer has regarded his L.primccvum as identical with such obcor- 

 date leaves of L. MeeMi, as the last mentioned three varieties. Dr. New- 

 berry remarks {I.e., p. 4), that "indeed the probabilities are against it, 

 since no intermediate forms have been found, and none of the panduri- 

 form leaves of L. MeeMi have been obtained from Greenland, where 

 obovate, entire or emarginate leaves similar to those given the above 

 names {Phyllites obcordatus and Leguminosites Marcouanus) do occur, 

 and also many of the emarginate, oblong-ovoid, or lanceolate leaves, 

 which I have called Liriodendron simplex.'''' 



It is now to be remembered, as mentioned before, that such obcordate 

 leaves may occur, and indeed do commonly occur, together with leaves 

 of the typical form on the same tree. I consider it rather as acci- 

 dental that none of the pauduriform leaves of L. MeeMi have yet been 

 discovered in Greenland, and it is strange that Doctor Newberry does 

 not see any gradation forms in the Greenlandish leaves of L. MeeMi, 

 figured by Heer on Plate xxii. Figs. 12 and 13, and farther on Plate 

 XVIII, Fig. 4c (?. c). 



Massalongo is another author with the same view in regard to the 

 variation or rather the occurrence of different forms of a fossil imo- 

 dendron s\)ecies. We see in the Plates of his "Flora Senigalliese" * 

 a series of leaves of a species which he has referi ed to L. Procaccinii 

 Ung. This species, which was first described by Unger,t is character- 

 ized as follows: ^^Liriodendron foliis trilobis, lobo medio maximo trun- 

 cato — emarginato, lobis lateralibus obtusis vel acutis integerriinis, uervis 



*Massalougo: Flora Fossile del Seuigalliesc, 1859, p. 1511. 



tFr. Uuger: Genera et species plantarnm fossiliiiiu, 1850, p. 443. 



