eS 
442 
growth, as many other now treeless parts of 
North America have been, not excepting at 
least a large part of the present arid region; 
but those forests existed in other geological 
epochs, and they have been destroyed by sub- 
sequent unfavorable physical changes. The 
region of the great prairies has also been shorn 
of its forests once, and perhaps twice, since the 
tertiary period: that is, in the tertiary period, 
and even before, an extensive arboreal flora 
prevailed in North America, which was closely 
related to that which now exists; but, with 
the accession of the glacial epoch, the forests 
of the region here discussed were necessarily 
wholly destroyed, except, perhaps, along its 
southern borders. 
Accumulating evidence seems to show, that 
there was an interglacial epoch of temperate 
climate, during which that great region was 
again covered with forests, and that these 
were in turn destroyed by the second glacial 
epoch. It is the remains of these interglacial 
forests that have been so frequently found in 
excavations made in the prairie region, and 
which have excited so much local interest. 
Those forests were evidently extensive ; but, 
unlike those now living there, they seem to 
have consisted largely of conifers.. I do not 
doubt, that, at the close of the second glacial 
epoch, the present prairie region of the United 
States was as completely destitute of trees as 
any of its present prairies were when white 
men first discovered them. ‘The opinion also 
seems a reasonable one, that the foresting of 
the prairies has been slowly in progress, ever 
since the close of the second glacial epoch, by 
the process of natural dispersion, and, further- 
more, that this dispersion of trees progressed 
mainly from the south-eastward. Not only 
were the interglacial forests necessarily de- 
stroyed by the icy visitation of the second 
glacial epoch, but the whole, or nearly the 
whole, surface was rewrought, and practically 
a new soil was produced by the glacial action 
and the subsequent physical conditions. 
Such a new soil would naturally be first oc- 
cupied by herbaceous plants, whose abundant 
and annually matured seeds are so readily 
distributed by natural means. So, also, the 
pioneer occupants of the new land among the 
trees would doubtless be those whose light and 
abundant seeds are capable of being distributed 
by the winds, and whose most congenial habi- 
tat is upon the moist grounds which border the 
streams : such are the cottonwood, willows, and 
elm, forexample. It is especially the first two 
that are found to be the most advanced of the 
western arboreal pioneers upon the borders of 
SCIENCE. 
region. 
[Vor. IIL, No. 62. — 
the great plains, and which were doubtless the 
pioneers in the primitive foresting of the prairie 
Other trees followed those pioneers 
more slowly, for their methods of propagation 
were slower; but still the methods of natural 
propagation of the majority are sufficiently vig- 
orous to suggest, that, if the prairie fires had 
never been introduced, the early settlers would 
have found that great region a forested instead 
of a prairie one. 
How long the battle of the fires against the 
trees continued is not known; and by what 
successive steps the latter succeeded in gaining 
and holding even the small strips of land along 
the borders of the,streams of so wide a region, 
hundreds of miles from the place of their 
original departure, it is difficult to say. It is 
probable that the pioneer trees effected their 
occupancy there, to a large extent, before the 
fires prevailed, and that their presence favor- 
ably modified the immediate conditions for the 
occupancy of other trees. The streams also 
seem to have favored their occupancy, not only 
by the additional moisture which they gave to 
the adjacent soil, but by acting as checks to the 
fires which alternately swept the prairies on 
each hand, lessening the average frequency 
of fires upon their bordering bottom-lands by 
perhaps one-half of what it otherwise would 
have been. 
The subject, as I have attempted to present 
it, may be summed up briefly as follows: in 
the natural geographical distribution of faunas 
and floras, nature necessarily fixes the potential 
boundary of such distribution at a greater or 
less distance in advance of the boundary of 
actual occupancy ; and, when these two bound- 
aries come to coincide, there is necessarily an 
end to distribution. When the prairie region 
was first known, the potential boundary-line 
of forest occupation was at least five hundred 
miles westward from that of full occupancy. 
At the close of the glacial epoch the whole 
of the great prairie region was practically des- 
titute of vegetation, but its new soil was capable 
of supporting an abundant and varied growth. 
Herbaceous vegetation first occupied the soil, 
and trees followed more slowly. The obstacles 
to the occupancy of the new soil by forest-trees 
at the close of the glacial epoch were, first, the 
slowness of the process of natural distribution ; 
second, the pre-occupancy of the soil by her-— 
baceous vegetation, preventing or retarding the © 
effective germination of the seeds of trees; — 
third, the subsequent prevalence of annual fires _ 
upon the grassy surfaces, which retarded forest- 
growth. 
The conditions favorable to the natural dis- 
