March 15, 1907] 



SCIENCE 



431 



branches of the deeply incised drainage (ex- 

 cept in the case of the very smallest examples), 

 have eroded deep valleys in the mica schist. 

 On the other hand, the Tallulah River is 

 found to cross a resistant barrier of dense, 

 hard quartzite near its junction vyith the Chat- 

 tooga. This rock is wholly distinct from the 

 mica schist, contains little else than fine 

 quartz grains firmly cemented into a massive 

 bluish-gray quartzite. It is less apt to be 

 sheared than the other rocks of the region, 

 having to a great degree withstood the crush- 

 ing incident to regional metamorphism ; it 

 weathers with difficulty, and wherever present 

 forms a serious obstacle to stream erosion, 

 even a small layer of it determining the site 

 of falls on some of the branch streams. A 

 great thickness of this resistant rock is crossed 

 by the lower course of the Tallulah. A tra- 

 verse of several miles along the lower course 

 of the Chattooga resulted in the finding of 

 nothing but mica schist in the ledges and in 

 the boulders in the stream bed, so it seems 

 apparent that this stream has not encountered 

 any of this hard rock in its down-cutting, but 

 is located wholly on the weaker mica schist. 



The presence of the falls in the Tallulah 

 Eiver and their absence in the Chattooga now 

 become perfectly intelligible. When the cap- 

 ture occurred falls were produced which at 

 once began to be worn back. After retreating 

 to the junction of the Tallulah and Chattooga 

 Elvers two series of falls were formed, one 

 retreating up the course of the former river, 

 and one up the course of the latter. The falls 

 in the Chattooga were worn back to grade 

 during the long period of time which has 

 elapsed since the capture, for the weak mica 

 schist ofPered no great obstacle to the river's 

 attempt to grade its course. The falls in the 

 Tallulah have not been thus worn back, since 

 the hard-rock barrier with which that stream 

 had to contend has made the process of 

 grading a very slow one, a retreat of but a 

 few miles having been accomplished thus far, 

 and the work still continuing very slowly 

 under the extremely unfavorable conditions 

 which prevail. 



There are many direct evidences of river 

 capture which have been obliterated during 



the long period of erosion since the capture 

 occurred. The former channel of the Chat- 

 tooga across the divide to the southwest 

 cannot be seen; stream dissection has been 

 extensive and all traces of that channel itself 

 have been removed. The valley of the Chat- 

 tooga has been so widened that no bench or 

 terrace remains to indicate the abrupt change 

 from the former mature upland valley to the 

 deeply incised gorge. The lack of adjustment 

 between the shrunken beheaded stream and 

 the broad valley made to accommodate the 

 former larger river, has given place to a newly 

 established adjustment. The evidences of 

 recent capture are lacking, and everything 

 points to a great lapse of time since the cap- 

 ture, permitting the effacement of the more 

 temporary effects of capture, and a consider- 

 able degree of adjustment to the new order 

 of things. The evidences of remote capture, 

 however, are none the less conclusive. 



The theory of capture has been supported 

 on the basis of certain facts in the distribu- 

 tion of the fresh-water faunas. It appears 

 that a few shells from the Chattahoochee and 

 more western drainage basins are found in the 

 Savannah River, while a number of forms 

 from the Savannah system are found in the 

 Chattahoochee system. It is argued that these 

 forms must have passed between the two sys- 

 tems at the time of capture. 



That the fauna of the upper Chattahoochee 

 (the Chattooga River) might be transferred 

 into the Savannah River by the capture would 

 appear quite possible. It is much more diffi- 

 cult to account for the transfer in the opposite 

 direction, however, if we limit ourselves to 

 river capture as the means. Yet the main 

 transfer is supposed to have been in that 

 direction. That shells could have passed from 

 the lower level of the Tugaloo up 500 or 600 

 feet over falls and rapids to the higher level 

 of the Chattahoochee, does not seem probable. 

 It is possible, of course, that the headwater 

 portion of the capturing stream may have 

 been a more even slope instead of a series of 

 falls, but the evidence of other streams work- 

 ing headward into the escarpment suggests 

 that the capture was most probably initiated 

 by a series of more or less prominent cata- 



