18 LEAVES OF LIRIODENDKON HOLM. 



and sbow a considerable diflference from tbe mature leaves. A drawing 

 of one of the cotyledons is given on Plate iv, Fig. 1, and a germinat- 

 ing plantlet has been figured by Mirbel in his " ISTouvelles recherches 

 sur les caracteres anatomiques et physiologiques qui distinguent les 

 plantes monocotyledoues des plautes dicotyledoues."* But this cir- 

 cumstance that there is a difference between the cotyledons and the 

 mature leaves does not seem to be of particular importance, when we are 

 looking for the primeval leaf of any genus. We know well enough 

 from the numerous contributions on the subject of the germination, 

 how widely different the shape of the cotyledons can be from that of 

 the mature leaves of the same plant, and 1 will merely call attention 

 to the lobed cotyledons of Tilia, the entire ones of Acer, the Umbelli- 

 ferm, etc. 



As to the nervation of the mature leaf of the recent Liriodendron, 

 this seems rather to show a four-lobed than a three-lobed leaf, or per- 

 haps it would be better to say a pinnately-lobed than a palmately-lobed. 

 The three-lobed leaf, for instance, of Sassafras, so exactly figured in Pro- 

 fessor Ward's paper : " The paleontologic history of the genus Platamis,^^ 

 must probably always be considered as a really palmately- veined and 

 lobed leaf, with the iwo lateral ribs of the lobes strongly developed and 

 proceeding from the base of the midrib, in pairs. But we quite often 

 find, on the other hand, small leaves of the genus Quercus with only one 

 lateral lobe on each side, but such a leaf would never on that account 

 be considered as three-lobed. And I propose to compare the leaf of 

 Liriodendron with a lobed leaf of Quercus, on account of the nerva- 

 tion as shown on Plate iv, where several forms are figured of full- 

 grown leaves. We see on these leaves that the lateral ribs almost 

 proceed in pairs to the lobes, but not at all from the base of the midrib. 

 The leaf is really feather- veined, therefore pinnately-lobed, when lobes 

 are present. Now as to the varieties of Liriodendron Tulipifera, but 

 few have been mentioned and described, namely: '■'■ a acutiloba : lobis 

 acutis acumiuatisque, /5 obtusiloha: lobis rotundatoobtusissimis," both 

 of A. Michaux, and finally a variety chinensis Hemsl. The two varie- 

 ties, described by Mi haux, have been accepted by Pursh, De Candolle 

 and Browne, though by the last named with a difference in the nomen- 

 clature (''rtcu^^/b/^a" instead of '■^acutiloba''''), and with Loudon as au- 

 thor, but the description agrees perfectly with that given by Michaux. 

 De Gaudolle has another variety " foliis quadri-lobis ant rarius subinte- 

 gris ovatis apice truucato-emarginatis=Z/. integrijolium hortul.," but 

 this variety has been considered as identical with Michaux's obtusiloba 

 by Loudon. Linne also made a variety " (5. Tulipifera caroliniana foliis 

 productioribus magis augulosis," which possibly may be the same as 

 Michaux's variety acutiloba. 



As to the variety chinensis, this was discovered in China by Dr. 



* Aunales du Museum cl'lust. uat., xiii, 1809, PI. VI. 



t Proceediugs U. S. National Museum, Vol. xi, 1888, PL xxi. 



