36 



if the lobes were altered as indicated by the dotted lines, would 

 greatly resemble Liriodendron giganteum Lesq., particularly its 

 variety cruciforme, in form, venation, and size. 



Fig. 9 shows a modern leaf which is identical with that of 

 Liriodendron Mcekii Heer, and I find many modern leaves simu- 

 lating this form more or less closely ; as a rule, however, the 

 lateral lobes are more oblique than in the form figured. 



Figs, i and 3 show modern leaves which are identical with the 

 Phyllitcs obcordatus of Heer, and which serve in a measure to 

 confirm the reference of this species (of Phyllites) to Liriodendron 

 primaevum. They also strengthen our conviction that Lirio- 

 dendron primaevnm Newb., Liriodendropsis simplex Newb., and 

 Liriodendropsis angustifolia Newb. are valid species of Lirio- 

 dendron, notwithstanding the fact that this view is criticized in 

 some quarters. 



Fig. 2 shows a modern leaf which has reverted to a still earlier 

 stage in the history of the genus, the stage in which the simple 

 ovate leaf had not yet become emarginate at the apex. 



Fig. 5 shows a leaf of peculiar form related in a general way 

 to the form figured at 12 as resembling Cissites acuminatus, but 

 narrower, with a slightly emarginate apex and rounded lobes, the 

 lateral margins and the primaries being somewhat more ascending. 

 Finally, at Fig. 17, we picture the reduced, two-lobed, long- 

 petioled, Liriophyllum-Yike. leaf which is often found on the tulip- 

 tree where there has been some diminution of nourishment, such 

 as is caused by proximity to flowers or among leaves developed 

 from forced buds. While this form is not constant in such situa- 

 tions, it is fairly common, there being an ever-present tendency to 

 produce leaves of this shape or approximating it. The bud-scale 

 of the blossom often bears at its apex a true leaf-blade almost 

 exactly similar to the one here figured. A number of these 

 leaves are shown in the September (1901) number of Torreya.* 

 They show that whereas a bud-scale has always been considered 

 the morphological equivalent of a leaf, in this genus they are 

 morphologically stipules, or modified leaf-segments. 



* Berry, E. W. Notes on Li?-iodendron leaves. Torreya, i \ 105-107, //. /, 2. 

 1901. 



