216 The American Geologist. October, 1894 
abandoning the pass through the range, we may have pre- 
sented the phenomenon of two streams in headwater opposi- 
tion with identical species; but the "wind gap" will remain 
as a monument of the time when the streams were part of the 
same system. 
Th ro ugh a llu v in t io n . 
Changes of less significance, but often throwing light on 
the dispersion and consequent modification of flu via tile life, 
occur in the deltas and alluvial plains of river basins. By 
increasing alluviation in the delta region, the tributaries of 
the main stream are often forced aside to flow in the lower, 
back levels of the plain, along which they may find their way 
to the sea many miles distant from the vagarious mouth of 
the master river. In this manner a fauna once common to the 
main stream and its tributary may be divided except for the 
occasional communication established by floods in the delta 
district. But if the separation thus initiated be continued by 
favorable geological conditions, the isolation of the two fau- 
nas may be so complete that any subsequent intermigration 
between the two river systems must take place through the 
salt water along the coast, or by one of the interstream 
changes already described. The Red river of Louisiana is 
now in this initiatory stage of separation from the Missis- 
sippi river. 
Through slight submergence. 
Where submergence takes place, the lower courses (if rivers 
become converted into arms of the sea. Drowned river valleys 
of this origin now abound in the northern hemisphere. Where 
this has occurred the freshwater fauna common to the master 
stream and its lower tributaries may be cut off from each 
other by the invasion of salt water, as in the case of those 
rivers which discharge into the present Chesapeake ha}'. 
Through elevation and revival of streams: 
Opposed to the preceding action on organic life, is the up- 
lift of a region like Hudson's bay, whereby the streams would 
be made to continue their courses seaward and frequently be- 
come confluent, thus bringing into contact species before liv- 
ing in separate basins. Whereas, in the case of submergence, 
the freshwater species are crowded back into a constantly 
lessening domain and thus are affected unfavorably, in this 
