432 



SCIENCE 



[N. S. Vol. XXV. No. 63T 



racts. These would effectually prevent trans- 

 fer through the water itself. Furthermore, 

 there is reason to believe that the transfer 

 of water may have been nearly or quite com- 

 pleted before an actual surface valley con- 

 nection was formed, since (as Lane has 

 pointed out) leakage through the rocks from 

 the higher to the lower level will go on for a 

 long time, possibly increasing until all of the 

 water from the higher level passes under- 

 ground to the lower stream, leaving a dry 

 channel for some distance below the point of 

 capture in the upper valley. Under these cir- 

 cumstances it would be difficult to conceive 

 of any transfer of faunas from the lower to 

 the higher stream which was dependent upon 

 direct fresh-water communication. On the 

 other hand, there are so many means for the 

 dispersal of fresh-water shells, and the evi- 

 dence in other localities is so conclusive that 

 they have been dispersed by such means, that 

 we may reasonably suppose shells from either 

 of the two systems might be transferred to 

 the other independently of the capture. For 

 this reason I do not believe that the distribu- 

 tion of the shells can be urged as a proof of 

 capture, although the fact of capture is well 

 attested by other lines of evidence. 



The conclusion in favor of the theory of 

 capture is further confirmed by the presence 

 of old river gravels along the former south- 

 westward course of the Chattooga. In view 

 of the fact that the capture occurred at a 

 remote period, we should not expect to find the 

 former channel preserved, nor to find the 

 gravels deposited by the river along that 

 channel in their proper place. The stream 

 dissection which we have already seen de- 

 stroyed all traces of the former channel, 

 would also wash the gravels down the slopes 

 of the growing ravines and valleys, in many 

 cases removing them altogether, but possibly 

 leaving remnants in specially favored spots 

 on the slopes and in the valleys. A careful 

 search revealed the presence of these gravels, 

 usually as scattered pebbles and boulders on 

 the hillsides, but occasionally as considerable 

 patches in the bottoms of small branch val- 

 leys. There was no mistaking their char- 

 acter. The present stream-borne material, 



even in the largest of these branches, is quite 

 angular and evidently of local origin. The 

 gravels are beautifully rounded quartz pebbles, 

 cobbles, and boulders, somewhat roughened 

 where exposed to the weather for a long time, 

 but perfectly smooth where recently unearthed. 

 In one place they were so abundant that a 

 farmer had made numerous large piles of 

 them in an ineffectual attempt to clear a 

 small plot of ground for agricultural purposes. 

 Their occurrence, together with the unequivo- 

 cal topographic evidence, would seem to re- 

 move the question of capture from the realm 

 of theory, and place it definitely in the realm 

 of known facts. 



Douglas Wilson Johnson 

 Cambridge, Mass. 



reasons for believing in an ether 



Many scientific men who are not physicists 

 feel an ill-defined distrust of some of the 

 more or less complex conceptions of modern 

 physics. They feel that the physicist has, per- 

 haps, allowed his imagination to carry him too 

 far and has not stopped often enough to re- 

 examine the foundations of his faith. 



Perhaps the most fundamental conception 

 exciting some such distrust from the outside 

 is that of the aether which is assumed to fill 

 all space. The non-physicist who has read of 

 the oft-repeated but entirely unsuccessful at- 

 tempts to detect the ' ether wind ' due to the 

 earth's movement through space, and of the 

 negative results of all ' direct ' experiments on 

 the tether, begins to feel that the builders of 

 physical theory are perhaps unreasonably 

 tenacious of an idea which could, perhaps, 

 best be dispensed with. 



It may not be out of place, therefore, to 

 state as briefly and clearly as possible several 

 reasons for belief in an ether, reasons suf- 

 ficient because based directly on observation 

 or experiment. 



The most important evidence is the simple 

 fact that ihe velocity of light does not depend 

 on the velocity of the source. This is shown 

 by the normal apparent shape of the orbits of 

 binary stars, which, it is easy to see, would 

 otherwise appear distorted. For, if the orbital 

 velocity of one member of a binary star 



