318 ROYAL SOCIETY OF CANADA 
10 and 22), and Siberia (24), from, which it would appear that as we 
now know it, this genus is essentially distinctive of the Eocene age. 
Phragmites is a form of plant remains which is nowhere clearly 
defined, though in a general way it may be recognized without much 
doubt. Precisely what it embraces with respect to either genus or 
species, it would be impossible to say at present, though in a general 
way it may be said to embrace fragments of broad leaves, more rarely 
fragments of stems or even of rhizomes of Monocotyledonous plants. 
The fragments of leaves are not always separable with certainty from 
other Monocotyledonous leaves with similar characteristics, while the 
stem fragments are clearly differentiated from Cyperacites. The rhi- 
zomes are usually sufficiently well characterized to be recognized with 
certainty. There is no correlation between these various forms rele- 
gated to a common genus, but when recognizable their characters are 
sufficiently definite to permit of using them for stratigraphical pur- 
poses. An examination of the North American distribution of the 
genus shows a somewhat wide range. Thus, P. cretaceous, Lesq., 
represented by both leaves and rhizomes, is a constituent of the flora 
of the Dakota Group (42: p. 34, and 43: p. 37). Dawson has reported 
the leaf of P. cordaiformis, Dn., from the Upper Cretaceous of Van- 
couver Island (9: p. 26). Newberry reports fragments of leaves of 
an undefined species from the Cretaceous (46: p. 27, pl. xxii, f. 5), 
and Ward, in his Synopsis of the Laramie Flora enumerates four species 
as belonging to the Laramie proper, with two from the Senonian (56: 
pl. xxxil). On the other hand, Lesquereux reports one species from 
the Tertiary (42: p. 141), and Knowlton (35: p. 779) reports P. 
latissima from the Fort Union Group. Reference to Heer’s well-known 
works (29, 26, 24, 23, and 31: p. 161) shows four species confined to 
the Eocene of Europe and Greenland, of which P. oeningensis, A. Br., 
is by far the most frequently represented. This summary shows nine 
Cretaceous localities against six Tertiary, and as these latter are all 
Eocene, it is clear that while Phragmites is common to the Upper Cre- 
taceous and Lower Eocene, it is more typical of the former than the 
latter. | 
Reviewing the facts thus dealt with, we can only conclude that the 
flora of the Kettle river is certainly not Cretaceous, and that in its 
general facies it is Eocene rather than Miocene. This conclusion, how- 
ever, necessarily raises an important question as to the particular age 
of floras previously determined and provisionally referred to the Mio- 
cene (51: iv, 68 and 52; iv, 36, etc.), and especially with reference to 
a critical comparison with the Similkameen flora as already determined 
