THE FLORA OF THE SAINT EUGENE SILTS 



407 



mediate proximity to the petiole; margins entire, lobed, or 

 lobate-dentate ; nervation 3-palmate from the base ; lateral pri- 

 maries branched from the onter sides, a basilar branch occasion- 

 ally simnlatiiig- a weak, exterior lateral primary, the branches all 

 coalescing and becoming camptodrome in the marginal regions. 



The type of foliage represented by these specimens is more 

 abundant than any other in the collections from the Saint 

 Eugene silts. Almost every piece of matrix contains remains 

 more or less complete or fragmentary. Individual leaves differ 

 so widely between themselves, in many instances, that the ques- 

 tion of mutual specific identity might well be raised and, if found 

 dissociated, either stratigraphically or geographically, they 

 would probably be regarded as specifically distinct. In the silts, 

 however, they are found not only closely associated but also, as- 

 sociated with them, are other specimens of leaves that show 

 every possible gradation of form between the extremes, so that 

 if a considerable series is arranged it is impossible to draw any 

 satisfactory line or lines of demarkation between them. 



Among existing genera in which heterophylly is a con- 

 spicuous feature may be noted certain ones in the Menisper- 

 maceae, especially Cissampelos, Menispermum, and Cehatha 

 (Coccul'us) ; and it is, in many instances, difficult to differentiate 

 between these genera by means of the leaves alone. Comparison 

 of our specimens, however, with a large number of leaves of spe- 

 cies in these genera, particularly with those of the existing Ce- 

 batha Carolina (Linnaeus) Britton, ajDpeared to indicate a some- 

 what closer resemblance to the latter. This was my original 

 generic identification {loc. cit.), based upon the specimens repre- 

 sented by FIGURES 1-3, PLATE 39, and it may be regarded as more 

 or less significant that Berry'* described and figured specimens 

 from the Pleistocene of Kentucky, that closely resemble ours, 

 which he identified as Cehatha Carolina without question. Our 

 specimens are larger in size ; but otherwise it would be difficult 

 to ditferentiate between them. 



28 Berry, E. W. The Mississippi River bluffs at Columbus and Hickman, Ken- 

 tucky, and their fossil flora. U. S. Natl. Mus., Proc. vol. 48, no. 2074, p. 300, pi. 12, 

 figs. 3-5. 1915. 



