246 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
lah alone, but does not make clear what they could be. It is difficult 
to imagine movements so profound as to produce a splendid series 
of cataracts in a deep, narrow gorge, and yet so limited areally as to 
have no effect on any of the other streams close by, and so obscure as 
to defy detection. 
Finally, it may be noted that the distribution of the river gravels 
is inexplicable on the basis of any other theory than that of capture; 
and that this latter theory affords full and reasonable explanation 
for all the features noted in the district. 
Summary. 
After a brief review of the literature on the Tallulah district, and 
a statement of the problem of river capture there presented, we have 
considered the relations of the region to the greater geologic and 
physiographic provinces of the southeastern United States. It was 
shown that the region lies wholly within the crystalline belt, on the 
western edge of the Piedmont plateau and the eastern edge of the 
southern extension of the Appalachian mountain belt. In the more 
detailed study of the local geology we found a series of mica schists 
and gneisses, striking northeast-southwest and dipping southeast, 
containing occasional beds of slate, limestone, and quartzite, and cut 
by pegmatite veins and dykes of diabase and diorite. The most 
striking features of the local topography were found to be the two 
peneplain levels separated by a marked escarpment, the open valleys 
on the upper level and the deep gorges cut far below that level, the 
hanging valleys of minor branch streams, the extensive falls on the 
Tallulah River, and the elbow bend of the Chattooga-Tugaloo. The 
process of river capture was then- analyzed, and a series of expectable 
results deduced. When compared with the actual features in the 
Tallulah district the theoretical results of remote capture were seen 
to have a close correspondence to the observed facts, which furnishes 
a strong presumption in favor of the correctness of the capture theory. 
The distribution of river gravels completes the evidence in favor of 
capture, whereas the alternative theories advanced, fail to account 
for this and other features observed. While the distribution of 
freshwater shells does not appear valid evidence of capture, the 
other evidence is so convincing that we conclude the former head¬ 
waters of the Chattahoochee have been diverted to the Savannah 
