INTRODUCTION. 5 



a single leaf, or species, characters which are now generally found separated 

 in far distant families of plants. The leaves of Eremophyllum, so striking 

 by the peculiar appendages of their borders ; those of Anomophyllum, refer- 

 able to Platanus by one-half, to Quercus by the other; those of Platanus 

 ohtusiloba, half Acer, etc., are of this kind. 



On another side, the characters of some of the Cretaceous species are 

 sometimes of such a transient or indefinite order that it is scarcely possible 

 to take hold of them and to describe them with any degree of reliance. At 

 first sight they appear very distinct, but, in comparing a number of spec- 

 imens, the differences dwindle by unmistakable transitions and disappear. 

 In other leaves, on the contrary, visibly identical by their outlines, the 

 nervation is so different that they are forcibly separated and referred to 

 far distant generic divisions. Hence this flora does not leave any satis- 

 faction, any rest, to the mind. Even the most clearly defined types become 

 doubtful in regard to their integrity when we see others, which, at first, 

 were recognized as positively fixed, manifesting instability and pointing to 

 diversity of relation by the discovery of new specimens. The leaves con- 

 sidered first as Sassafras, for example, seemed evidently referable to this 

 genus; but when leaves of the same type were found with dentate borders, 

 though bearing, besides, all the characters of a genus which belongs to the 

 Lanrineoe, a family where, as yet, no representative has been found with 

 dentate borders of leaves; when others were obtained with subdivisions 

 of the lower lobes in two or three, thus showing the palmate shape of 

 Aralia leaves, the confidence in the value of the characters at first recog- 

 nized had to be abandoned. 



The first exposition of the Dakota Group flora shows four species of 

 Ferns, six species of Conifers, and one of Cycadeoe only. To this small 

 number we have added in this volume one species of Gleichenia, six species 

 of Conifers, and five of Cycadece. The specific values of some of the vege- 

 table remains referable to the Conifers is, however, doubtful, especially 

 for those which are represented by cones only. AUetites Ernestince, 

 Sequoia formosa, Sequoia BeichenhacM, and the fragments described as 

 Inolepis are of this kind; all, however, though their specific or generic 

 relation may be uncertain, are evidently representatives of some species of 



