Bciseleveling and Organic Evolution. — Woodworth. 215 
most mountain ranges which are continuous, and which consequently 
must from an early period have completely prevented the inosculation. 
of the river systems on the two sides, leads to the same conclusion.* 
Through headwater diversion. 
Darwin had in mind evidently those changes which a care- 
ful study of rivers has established as a part of their history. 
It is unnecessary here to refer to particular instances} in 
which by changes in the headwaters of a river, some portion 
of its drainage has been diverted into another system, carry- 
ing with it of course a representation of its fauna and a 
change of condition in the whole life of both streams. It has 
been further shown that, at successive times, the headwaters 
of a river system have been diverted into a contiguous basin, 
so that, since the changes are slow and gradually accom- 
plished, each new capture may introduce into the fauna of a 
river basin new varieties, which would find themselves in 
close contest with allied species already introduced, as well as 
with the older species peculiar to the river system. Dispersal 
may in this manner be accomplished across a main summit of 
land, such as that now dividing the waters of the Mississippi 
basin from those of the Atlantic slope, or it may be limited to 
interchanges of fauna between parallel streams occupying the 
same continental slope. 
Through antecedent streanw. 
The not altogether acceptable case of the Green river, 
which Hows through the Wahsatch range in Utah, may be 
taken as an example of rivers of this class, where the uplift 
of a mountain barrier has been successfully resisted by the 
river. So long as unsurmountable rapids or falls do not in- 
tervene, there is a free communication across the range. 
Even where falls of considerable magnitude occur, certain 
fishes have acquired the habit of leaping them in their at- 
tempt to reach the headwaters during the spawning season. 
If, at a late date- in the history of an antecedent river, its 
headwaters be tapped and led to the sea in another direction, 
*Origin of Species, sixth (Am.) ed., New York, p. 344. 
IVY. M. Davis: A River Pirate, Science, mi. 1889, p. ins. 
R. DeC. Ward: Another Riser Pirate, Science. \i\. 1891, p. 7. 
H. Li. Harris: A new instance of stream capture. Science, xxir, 1893», 
pp. :'-<;. .T7. 
Collier Cobb: A recapture from a River Pirat«, Science, wit. 1893, p. 
195. 
