124 
Phragmites, have been described from Cretaceous and even older 
horizons, but fossils which can be unquestionably referred to the 
grasses can hardly be said to date back beyond the Eocene 
Tertiary.* 
The figures with which ours may be most closely compared 
are of species from either Eocene or Miocene strata in the Old 
World or Greenland, included under the genera Phragmites + and 
Arundo.t 1 do not know of any having been heretofore described 
from the eastern United States, although broad grass-like leaves, 
which I am inclined to refer to Phragmites, occur in the yellow 
gravel sandstone at Bridgeton, N. J., as previously noted by me.§ 
The flora of the Bridgeton sandstone is almost certainly 
Miocene or early Pliocene in age, and we may safely refer our 
species to about the same horizon. Itis greatly to be desired that 
leaves should be found in connection with the Staten Island speci- 
mens and that culms and rhizomes be found at Bridgeton. . Should 
such discoveries be made, a more exact comparison between the 
two floras would be possible. 
*For a general discussion of this subject see « Fossil Grasses,” J. Starkie Gardner, 
Proc. Geologists Assn. 9: No, 6, in which copious references may be found. 
+P. Oeningensis Al. Braun. 
Heer. Fl. Tert. Helvet. 1: 64. p/. 22, figs. 5a-e; 24; 27, fig. 2b; 29, fig. 8¢. 
Ludwig, Palzontog. 8: 80. p/. 16, figs. 1-10 ; 18, figs. 2-21; 24, fig. 7. 
Ettingsh. Fl. Bilin, 97 [21] pl. 4. figs. 6-70. 
$A. (Donax) Goepperti (Miinst.) Heer, Fl. Tert. Helvet. 1: 62. pi. 22. Sigs: 
2p Of ay 
Ettingsh. Fl. Bilin, 95 [19] A/ 4. figs. 2-4. 
Ludwig, Palzontog.8: 80. p/. 77. 
A, anomala (Brong.) Heer, Fl. Tert. Helvet. 1: 63. pl. 22, fig. 4. 
§Bull. Torrey Bot. Club, 19: 330-333. 
