said to have been found in or near this particular swamp, I did not 

 see. Magnolia glauca {M. virginiand) might have been expected 

 in such a place, but apparently it is not known from the western 

 hall of Long Island. 



Nearly half the shrubs in both habitats are of the Ericaceae 

 and closely allied families. The Compositae constitute nearly 

 half the herbs noted in the uplands and a considerable proportion 

 of those in the swamps ; but more complete lists, including spring 

 and summer flowers which become unrecognizable in the fall, 

 would doubtless show a smaller proportion of this family. Ever- 

 greens other than pines are very scarce in the dry pine-barrens 

 and not very numerous in the swamps. 



The most interesting features of this pine-barren vegetation are 

 brought out by a study of the geographical distribution of its 

 components. For the sake of brevity the species of both habi- 

 tats may be combined, since it happens that the geographical 

 affinities of both are much the same. 



Not one of the species whose identity is certain is confined to the 

 coastal plain, though Pieris Mariana, Clethra, Myrica, Ilex glabra, 

 and Eiipatoriitin verbenaefoliuin are mainly so.* On the other 

 hand, Arctostaphylos, Behila populifolia, and Panicularia canadensis 

 are chiefly confined to the glaciated region. Azalea viscosa, Leii- 

 cotho'e, Didichiiim, and Eriophoriim are widely distributed both in 

 the glaciated region and coastal plain, and also occur more rarely in 

 bogs in the southern Alleghanies.t Nearly all the species are 

 common in New England, especially in the southeastern portion.;}: 

 At least one-third of the dry pine-barren plants are common to 



* Nearly all the few endemic coastal plain species known from Long Island seem 

 to be bog plants, confined for some reason not fully understood to the eastern half of 

 the island. 



t See Rhodora 7: 72-76. 1905. At that time I supposed Diilichium to be 

 strictly confined to the glaciated region and coastal plain, but I have since noticed 

 that it has been reported from the mountains of Kentucky by Kearney (Bull. Torrey 

 Club 20 : 479. 1893) and from Cullman County, Alabama, by Mohr (Contr. U. S. 

 Nat. Herb. 6: 396. 1901). 



+ See HoUick, Bull. N. Y. Bot. Card. 2: 381-407. 1902; Blankinship, Rho- 

 dora 5 : 128, 129, 133. 1903. I have found most of them also in Worcester Co., 

 Mass., and some of the most abundant species extend at least as far up as northern 

 New Hampshire, according to Chittenden (U. S. Forestry Bull. 55 : 69, 99. 1905 ). 



