[ PENHALLOW] A REPORT ON FOSSIL PLANTS 3138 
This flora represents a remarkable preponderance of woods, several 
of which are entirely new, and in such cases previous knowledge cannot 
be utilized as an indication of the horizon they represent. There are, 
nevertheless, certain well defined forms of known value, and these will 
serve as a means of determining the horizon in connection with the 
general facies of the entire flora. From the list, we may nevertheless 
exclude the fern stipes, the exogenous wood and the undeterminable 
leaves as throwing no light whatever upon the problems before us. 
The genus Picea, although somewhat sparingly known in the fossil 
state, is, nevertheless, found to be distributed through a rather wide 
range of horizons. It is a well recognized and rather abundant element 
of the Pleistocene flora, in which several existing species are represented 
by both wood and leaves (48). Similarly also, Knowlton (38: p. 215) 
has shown that existing species are still undergoing deposition wherever 
local glaciation is in progress. Picea quilchenensis, Penh., has been 
recognized somewhat recently in the Oligocene of the Quilchena basin, 
British Columbia (1), but as the plant is represented solely by its 
leaves, it is impossible to determine its precise relation to other fossil 
forms, although the character of the foliage offers a suggestion that it 
may be related to the existing P. breweriana or P. sitchensis. More 
recently, Berry (3) has determined the extension of the genus into the 
Upper Cretaceous formation of New Jersey. In the Cliffwood clays he 
has found beautifully preserved cones which he regards as representing 
a species comparable with the existing P. excelsa. 
In 1904 Knowlton recorded the occurrence of a Picea in the Upper 
Eocene deposits at Kukak Bay, Alaska (33). This plant, which he 
designates as P. harrimani, is represented by its cones only, but these 
are in a fine state of preservation and permit of the inference that it 
most nearly resembles P. sitchensis among existing species. 
The present evidence shows our knowledge of Picea in the fossil 
state to be based upon the wood chiefly, though to some extent upon the 
cones as representing three species within the limits of the United States 
and ‘Alaska, but as these latter are in no way related species, they fur- 
nish no very precise basis for conclusions respecting the geological age 
of the horizon from which P. columbiensis has been derived. While 
our present limited knowledge of the genus leads us to believe that it 
should be looked for in the early Cretaceous at least, its present aspect 
is definitely Tertiary and chiefly Eocene. 
Cupressoxylon was first recognized by Penhallow (47) in the Cre- 
taceous formation of Medicine Hat, where the wood was found in 
abundance. Its more recent appearance in the deposits of the Kettle 
