508 Palaeontologie. 



Berry, Edward W, NotesonthePhy logen y o f Liriodendron. 

 (Bot. Gaz. Vol. XXXIV. 1902. p. 44.) 



The author reviews at some length, the chief arguments in 

 support of the idea that at least many of ihe fossil species of 

 Liriodendron are invalid, and in particular, cites the conclusions 

 reached by H o 1 m as based upon the variable nature of the 

 leaves in L. tulipifera. He also advances the chief reasons 

 for considering the various species as valid, in accordance 

 with the views of most American and European paleobotanists. 

 He agrees with Holm in regarding the primitive ancestral type 

 of Liriodendron to have been a simple, Magnolia-Wke leaf, since 

 all modern relatives of Liriodendron have such leaves, while 

 there is also a progressive simplification and reduction in loba- 

 tion as we proceed back in geological time, the most primitive 

 forms having ovate or oblong simple leaves. In the growth of 

 the modern Liriodendron tulipifera, a parallel development may 

 be observed, extending from the youngest, entire or merely 

 notched forms to those of the mature, typically lobed leaves. 

 He considers the primitive form of the entire, oblong leaf, 

 tapering at both ends, to have been succeeded by a series of 

 forms ranging from Liriodendropsis angustifolia Newb., through 

 Liriodendron simpiex Newb., L. primaevum, Newb. and L. 

 Meekii, Heer to L. semialatum of Lesquereux. This some- 

 what closely related group includes four types of leaves. The 

 first is narrow and elongated with an emarginate apex as in 

 L. angustifolia of the Amboy Glays. It is succeeded in 

 L. simpiex by similar but broader leaves and this form passes 

 by insensible gradations into elongated, emarginate forms with 

 a constantly increasing width ancl shoriening of length, while 

 the emarginate apex iinally becomes retuse. Here very slighi 

 alterations of form would render comparison with Liriodendron 

 tulipifera and Phylliies orbicularis possisible, as illustrated by 

 L. simpiex and L. primaevum. The fourth type of leaf is some- 

 what removec! from ihe other three. While retaining the an- 

 cestral form of the pointed apex, it becomes broadened and 

 develops an obtuse basal lobe on each side as found in 

 L. semialatum and often in L. tulipifera. The small leaves of 

 L. succedens Dn., from the Upper Cretaceous of Vancouver 

 island, are regarded as the natural descendants of L. semialatum, 

 hence it becomes possible to recognise in the theoretical, ob- 

 long-ianceolate type, stage 1, and in the five types noted 

 above, stage 2. 



Another series of types appears in Liriophyllum of the 

 Dakota Group, distinguished by their coriaceous texture, 

 obscure venation, deeply bilobate forms and their wide deviation 

 from the general type of Liriodendron with which it is difficult 

 to establish relationship, although it seems possible that a 

 shortening and widening of the leaf in Liriodendropsis angusti- 

 folia might lead to the development of transitional forms 

 passing through Liriodendron alatum to Liriophyllum. In Li- 



