478 MESOZOIC FLORAS OF UNITED STATES. 



some localities where the plants seem to have been abundant the rock 

 matter was not adapted to preserve them. In other cases the material 

 seems to have floated for a long time in water and hence it is much 

 macerated and comminuted. 



The abundance of plant impressions at a given locality can not 

 always be determined from the size of the collections made there. In 

 some cases the collection is the result of a single visit, in which only a 

 short time was spent in collecting, with no facilities for securing the 

 plants. In other cases repeated visits were made for the express purpose 

 of collecting. 



In the various collections some species are much more abundant 

 than others. This is not always due to the greater actual abundance 

 of these plants, for the accidents of preservation have had much to do 

 with the proportion of the different forms. But notwithstanding this, 

 the only way in which an idea of the importance of given plants in the 

 flora can be obtained is by noting the proportion of the fossils they have 

 left. For this reason, in giving an account of the different collections, 

 the number of specimens of the various species will be indicated by actual 

 count. 



LOCALITIES IN VIRGINIA. 



Professor Ward has made the following divisions of the Lower 

 Potomac of that State, taken in ascending order: (1) James River; 

 (2) Rappahannock; (3) Mount Vernon; (4) Aquia Creek. In the 

 writer's opinion the James River and Rappahannock are essentially the 

 same. Professor Ward's investigations show that the Aquia Creek beds 

 are only a few feet above the Mount Vernon strata and that there is 

 much in common in their flora. The Aquia Creek beds may be taken to 

 represent both. The principal difference in the flora is found in passing 

 from the Rappahannock into the Aquia Creek. 



These two horizons differ in some important points. The Rappa- 

 hannock horizon contains a large proportion of still surviving Jurassic 

 types. The dicot^dedons are very few in species and ver}^ rare in speci- 

 mens. They are, with hardly an exception, archaic in type, having a 

 vague, straggling nervation, with no marked differentiation in the devel- 

 opment of the nerves subordinate to the midrib. On the other hand, in 

 the Aquia Creek beds that occur at the Brooke locality we find compar- 



