520 THE AMERICAN NATURALIST. | (Vor. XXXIX. 
ization, and one occupying a higher position in the morphologi- 
cal scale, to occupy an inferior position in the geological scale, 
and thus to appear antecedent to morphologically inferior types. 
From the foregoing considerations it is clear that geological. 
data cannot be wholly relied upon to furnish a correct solution 
of questions bearing upon the evolution of plant forms, and this 
is especially true with respect to the broad-eaved angiosperms 
for reasons which will be described more fully on a subsequent 
page. Interesting as these various speculations may be, as 
applied to the Salicacex they nevertheless lack an essential 
foundation which can be obtained only by a more searching 
scrutiny of the family as a whole, and of its component genera, 
with respect to their more detailed distribution in Cretaceous 
and Tertiary time, as well as at present. 
Nowhere has it been possible to obtain so complete a record 
of the Cretaceous flora as from the exposures to be met with in 
Greenland and the United States, and the systematic way in 
which the various horizons have been worked out, affords a very 
satisfactory basis for a knowledge of the geological succession 
of species and genera. In his studies of the Cretaceous flora of 
Greenland, Heer found that out of a total of 88 species from 
the Kome beds, only 5.68% were Monocotyledons, while the 
Dicotyledons were represented solely by Populus primeva to 
the extent of 1.14 %. In the Cenomanian flora of the Atane 
beds, there was a remarkable increase of Dicotyledons amount- 
ing to 50.09 % ina total of 177 species; while from the Patoot 
beds of the Senonian, 59.48 % were Dicotyledons out of a total 
of 116 species, thus showing a marked development of this type 
of plants toward the close of the Cretaceous. This ratio be- 
tween the three divisions of the flora also appears to extend 
very largely to the Salicacez in particular, with respect to 
which we find 27.40 %, 42.40%, and 30.10 % respectively for the 
three horizons. An analysis of the family, however, shows that 
this relation does not altogether hold in detail for the individual 
genera. Thus in Salix the percentages are 41.40, 41.40, and 
17.20 which approximates to the results for the family as a 
whole with respect to the very evident falling off in the Upper 
Cretaceous. Nevertheless this is much more marked in the 
