ANALYSIS OF THE DAKuTA GROUP FLORA. 229 



Dakota Group, we have an indication of the grachinl march and develop- 

 ment of the vegetation, and are enabled to refer the origin of the dicotyled- 

 onous plants to the beginning of the Cretaceous period. For the leaves 

 of Populus recognized in the Kome schists, though the first observed fossil 

 remains of a dicotyledonous plant, ])robably do not represent the species of 

 that class of vegetation first produced. The active influences pro<lucing 

 gradual modifications must have existed for long periods before definite 

 results could be recognized by naturalists in the remains of a dicotyledonous 

 flora. We can not, therefore, expect to discover and recognize the first 

 representative of the new race, that of the Dicotyledons; but some valuable 

 conclusions on the nature and subsequent distribution of this new vegetable 

 group may be derived from studying the peculiar character of some of the 

 leaves of the C^enomanian. In looking over the lea\es of the Dakota Group, 

 which, in a flora of 460 species, represent 429 Dicot-s ledons, one can but 

 wonder at the work of natm-e which, in apparently so short a period, has 

 produced such an immense diversity of specific forms of leaves. The word 

 "specific" may seem hazardous. But it will suffice to examine the character 

 of a few of the leaves of the Dakota Group, to recognize not merely their 

 prodigious disposition to vary, but, at the same time, to effect such great 

 modifications in character that the result of the variations has often to be 

 admitted as implying- not merely specific but generic differences as well. 



ConsideiTiig the leaves of Liriodendron, for example, we find them 

 entire, ovate or oblong, always truncate or emarginate at the apex. In 

 Liriodcndroi/ prhiufriim Newb. (PI. XXVI, Figs. 1-4), with its svnonvms, 

 Legitmlnosiics MaycoHiniKti Heer and PhyUUes ohcordatus Hecr are consid- 

 ered by Heer as varieties of L. Ileekii. It is the true original form named 

 again L. simplex by Newberry, in bulletin of the Torrey Botanical Club, Jan- 

 uary, 1887. The base of the leaf enlarges, as shovsai in Fig. 2, but the specific 

 relation is still preserved. Gradually the leaves become more enlarged, 

 rounded, broadly, distinctly lobed on each side in the lower part, but remain- 

 ing ovate, obtuse or obtusely pointed, instead of emarginate at apex, clearly 

 representing a new species, L. semialatum (PI. XXV, Figs. 2-4; PI. XXIX, 

 Fig. 3). Then, as seen in PI. XXVIII, Figs. 5, )!, the leaves become con- 

 stricted in the middle, deeply emarginate at apex, and finally fiddle-shaped 

 or bilobate on each side in L. Meehii Heer. Still preserving a closely allied 

 form, but being greatly enlarged, with lobes at right angles or oblique, the 

 leaves represent the remarkable L. (jiffanieum, which is the Cretaceous type 

 of which the Tertiary, L. Proracrinii, and the living L. fidipifera are repre- 



