LONG ISLAND VEGETATION 



193 



barriers to the northward migration along the Atlantic coast are, 

 or have been, Chesapeake Bay, New York harbor, and eastern 

 ^Massachusetts. The first prevented the passage across it of 31.5 

 per cent of these plants, the second reduced the New Jersey 

 quota by 27.4 per cent, and the last formed an effectual barrier to 

 59.7 per cent of the plants that reach it. The large percentage of 

 species that fail to cross New York harbor points to greater bar- 

 riers in the past than at present. The dropping out of 8.3 per 

 cent of the species in central New Jersey also points to an early 

 postglacial barrier there. 



In the light of these data the following conclusions seem war- 

 ranted: (1) That we need not postulate an early postglacial land 

 connection to account for the pine barren flora of Long Island ; 

 (2) That the number of coastal plain plants on Long Island de- 

 mands a greater barrier than the present one to account for 

 its deficiency in these species; (3) That the present distribu- 

 tion of plants is in harmony with the geological evidence that 

 the land stood lower than at present in early postglacial times. 



For convenience of description the vegetation of Long Island 

 (Fig. 1) may be roughly divided into four interrelated types: (1) 

 the littoral vegetation, including the submerged marine algae, the 

 salt marshes, and the beach associations ; (2) the pine-barrens, 

 occupying most of the outwash plains east of Hicksville and -\mity- 

 ville and exhibiting two phases, the dry pine barrens and the 

 swamps; (3) the prairie of the Hempstead plains; and (4) the oak- 

 chestnut forests of the northern half of the island. 



There are today no data on the basis of which the plant forma- 

 tions of this region might be described in the order of their estab- 

 lishment following the retreat of the glaciers. The flora from 

 which the island derived its plant formations in early postglacial 

 times, under other climatic conditions, must have included many 

 boreal species long extinct. But a stud}' of the present associa- 

 tions furnishes information relative to the genetic development 

 of the vegetation at the present time, especially on nascent coastal 

 and denuded inland areas. Historical evidence has also been of 

 service in connection more especially with the forests and the 

 prairie. 



