May 20, 1892.] 



SCIENCE. 



below the confluence. The energy of westward headwater ero- 

 sion is unmistakable. All of the Loups bend that way at their 

 heads, and have their most vigorous tributaries on that side. 



It is noteworthy that the North Loup, having no large, aggres- 

 sive eastward neighbor, has retained its original head, the Cala- 

 mus. It has itself encroached upon the territory of the Middle 

 Loup, but that stream escaped capture by turning aggressor on its 

 own account. Possibly its original head was captured by the 

 North Loup. If so, it was after the Middle Loup had seized so 

 much territory westward, including the head of the large valley 

 in which Mud Creek now flows, that the conquest was a barren 

 one. It was no more serious in its effects upon the Middle Loup 



FIG. 1. 



Drainage map ol Central and Eastern Nebraska. The dotted lines along 

 the Middle and North Loups mark the present water-shed, lying in each case 

 nearer to the higher stream to the south-west. The short stream near x is 

 Lost Creek, so called because it disappears In the sands of the valley. The 

 line xy shows the trend of the buried cliffs of Cretaceous shales bounding the 

 old gorge of the Platte. 



than the capture of the Calamus by a tributary of Cedar Creek 

 would now be to the North Loup. 



Professor Todd has added some welcome and valuable contri- 

 butions to this discussion (Science, Mar. 11, p. 148), but his ob- 

 jection to the "efficiency of abstraction," on account of the 

 porosity of the strata in this region, does not appear to me to be 

 well taken for two reasons. First, the impression given by his 

 remarks, of the degree and extent of porosity, is exaggerated. It 

 chimes with a widespread popular notion of extensive subter- 

 ranean flows from one river to another, but the real exceptions to 

 that general hydrographic law which 'predicates the volume of 

 each river to be the product of rainfall on its own basin, are not 

 much more frequent or striking here than elsewhere. There is no 

 indication that the Blue receives any appreciable increment by 

 subflow from the Platte, or the Salt from the Blue, although both 

 are at a lower level than the larger stream to the west. Each has 

 a volume which may all be accounted for by the size of its basin and 

 the depth of annual precipitation. Those tributaries of the Salt 

 which approach nearest to the Blue are the weakest; if subflow 

 from the Blue were an important factor they should be the 

 strongest. The divide is formed by a moraine of the first glacial 

 epoch, running along the east bank of the Blue, where the words 

 " Big Blue" are written on the map. This moraine is the cause 

 of the peculiar arrangement of the tributaries of Salt Creek, and 

 of the abrupt turn to the south of all the Blue rivers to form the 

 Big Blue. It has the usual composition of a moraine — sand, 

 gravel, and clay. Many'examples of morainal lakes held up, and 

 rivers turned aside, by such material, testify to the fact that it is 

 not very porous. 



Secondly, headwater erosion would not cease on account of sub- 

 flow unless the latter absorbed the whole run-off. As long as 

 there are any surface streams, and they are rather numerous in 

 this region, they will erode their channels and, by virtue of the 

 law of unequal declivities, push the divide towards the higher 

 stream, ultimately abstracting the latter. If the subflow does 

 rob them of some part of their volume and eroding power, the 

 process will only be retarded, not prevented. 



I leave it to Professor Todd to answer the question of Professor 

 Davis respecting the deflection of rivers by rotation of the earth. 

 He has already adduced the Platte as an example, assuming it to 

 have flowed once in the channel of the main Loup. The accom- 

 panying profile (Fig. 3), reduced from one published by Chief 

 Engineer E. S. Nettleton (Irrigation Survey, U. S. Dept. of Agri- 



culture Progress Report, Part XL), will be useful in discussing this 

 assumption. The Loup at St. Paul is 95 feet below the Platte at 

 Grand Island. Since rivers do not shift from lower to higher 

 levels, it is physically impossible that the Platte should have 

 shifted from the Loup channel to its present position, unless there 

 has been a great change of levels. But such change is claimed. 

 Professor Todd thinks the Platte occupied the Loup channel (" the 

 north channel already described " cannot be other than that of the 

 main Loup, since it is said that "the Loups did formerly flow 

 through to the Platte " in that position) " when it was flowing on 

 a level seventy-five to a hundred feet higher, relatively, than at 

 pre.sent." Ths would bring it up to the position of the dotted line 

 O. C. Fig. 2, one hundred and ninety-five feet in the air above 

 the present Loup. There are no flood-marks, or other evidences, 

 to show that it ever flowed there. The " alluvian terrace," which 

 is the most significant and interesting feature of Professor Todd's 

 map, in its westward extension along the Loup, is obscured by a 

 range of sand hills, which form a broken and ragged divide be- 

 tween the Loup and Prairie Creek. Its main mass, aside from 

 the dunes blown up on its back, is below the present channel of 

 the Platte. It therefore furnishes no evidence that the Platte 

 ever flowed at a higher level between St. Paul and Grand Island. 

 On the contrary, it furnishes distinct evidence that the same rela- 

 tive levels, the same relative gradients (the Loup having less fall 

 than the Platte) and the same relative positions of the two streams 

 existed as far back as the second glacial epoch, substantially as 

 they now exist. Some obstruction at that time in the lower 

 Platte, possibly an ice-dam near Fremont, raised the waters till 

 they overflowed the divide at the head of Sand Creek. It is surpris- 

 ing that this new short cut did not become the permanent channel 

 of the Platte. Possibly the ice-dam extended below Ashland, but 

 with less elevation than at Fremont, thus permitting the new 

 channel to be cut down to its level, but not to the level of the old 

 channel. Hence the longer course by way of Fremont was re- 

 sumed when the ice retired. 



Both the Platte and the Loup are so heavily charged with sedi- 

 ment that a slight reduction of their gradients would cause depo- 

 sition of silt, and this result of retarded flow would be felt in both 

 streams far above the obstruction, but farther up the latter than 

 the former on account of its lower gradient. The ice-dam ponded 

 the Platte for some miles above ic, producing still water, in which 

 sediment rapidly accumulated. Thus was built up the eastern 

 end of the terrace to a level " seventy to ninety feet above the 



FIG 2. 



Proflle across the Loup, Prairie Creek, and Platte, running from St. Paul to 

 Grand Island, Neb. Horizontal line 1900 feet above sea-Ievel. i., Loup River, 

 1775 ft.; P.O., Prairie Creek, 1843 ft.; P., Platte River, 1870 ft. above sea-level; 

 O.C, alleged old channel of the Platte. 



Platte." Deposition, induced by retarded flow, an indirect result 

 of the obstruction, extended far up the Loup on account of its low 

 gradient. Professor Todd's map is correct in representing the 

 terrace as following the Loup instead of the Platte above their 

 confluence. From Columbus westward it is a Loup formation. 

 Not only does it draw away from the Platte, but it also sinks below 

 its level. It therefore furnishes no evidence of a change of levels, 

 or of a shifting of waterways, but rather of the persistence of 

 both as they now exist, since it fits well into present conditions. 



Another reason for doubting this alleged shifting of waterways 

 is found in the position and trend of the ancient rock trough of 

 the Platte. Its buried bluffs of Cretaceous shales have been just 

 touched by recent erosion sufficiently to reveal their existence 

 near the mouth of Beaver Creek, along Cedar Creek, and in the 

 bed of the main Loup between the mouths of these two creeks. 

 The trend of these ancient bluffs is shown by the line xy on the 

 map. It is oblique to the Loup channel, leaving the mouth of 

 Cedar Creek, and of all three of the Loup rivers, outside of the 

 Platte valley. If they ever entered the Platte directly and inde- 

 pendently, it must have been as indicated in my article of Jan. 



