^°m"'J PROCEEDINGS OF THE NATIONAL MUSEUM. 33 



II. Leaf apparently ovate without lobatiou. 

 L. laramiense Ward. 



L. Procaccinii Heer (PI. xxvi, Fig. 7 b, l. c.) 

 III. Leaf obcordate, more or less oblong. 



A. With rudimentary prolongation of the midrib. 



L. Meckii var. mucronulata Heer. 



B. Without any prolongation of the midrib. 



L. Mcekii var. Marcouana Heer. 

 L. Meekii var. primawa (Newb.) Heer. 

 L. Meekii var. obcordata Heer. 

 L. Procaccinii Uug. 

 L. simplex Newb. 

 (Leaf, of which the outlines are unknown: L. Hauerii Ettingsh.) 



There is, however, another character besides the above mentioned 

 and this is whether the base of the blade is horizontal or tapering into 

 the petiole, the last of which seems to be especially characteristic of 

 the obcordate leaves, but also varies. We find namely in L. Meekii, 

 var. mucronulata one leaf with the true obcordate shape, while another 

 one has the inferior margin of the blade distinctly horizontal, but in all 

 the other ones of Group in the blade tapers into the petiole. We 

 find the same kind of variation in the lobed leaves, as, for instance, the 

 varieties Helvetica and acutiloba o^ L. Procaccinii, and farther, the ynvi- 

 iit'ies 2^rim(cva and genuina of L. Meekii. while the base of the blade is 

 almost horizontal in the recent species. This fact does not seem to 

 speak in any absolute favor of a correct identification of the fossil 

 leaves with such a distinct tapering form of the blade into the petiole, and 

 as pointed out above there is good reason for consideriug some of the 

 obcordate leaves as belonging to plants of a quite different family, 

 namely, if we compare them with leaflets of the Leguminosa\ But we 

 must then see the articulation, and this very probably has been indis- 

 tinct, since there is no trace to be observed on any of the numerous 

 figures given of these fossil leaves. 



What form then, may be supposed as the original one of the Lirioden- 

 dro^ileaf ? I am incliued to think that we can conclude nothing posi- 

 tively in regard to the hitherto known fossil leaves identified as Liri- 

 odendron, and it is especially the presence of the rudiuientary pro- 

 longation of the midrib, observed not only in the recent species but 

 also in some of the fossil leaves, which gives us the greatest difficulty, 

 because we are entitled to conclude from this circumstance that the 

 leaf has not only been longer and with a larger number of lobes, but 

 also with an apex corresponding in size and shape to the lobes. There 

 IS, as it seems to me, the apparently entire leaf of L. laramiense and 

 some of the obcordate leaves of the other species, which are of the great- 

 est interest. We have but to elongate the midrib of these a little and 

 we will obtain either an obtuse or acuminate leaf, corresponding to the 

 genus Magnolia, and I myself am not unwilling to suppose such a form 

 as proper to the ancient type of the Tulip-tree. The identification of 

 Proc. I^. M. 90 3 



