PUBLICATIONS. 367 
If the gradient of the fjords is much greater than that of land val- 
leys, which most closely correspond to them in width and depth, the 
conclusion that the land “stood as high as the fjords are deep” would 
seem not to be warranted. Moreover, if the subsidence along normals 
to the coast was equal, the gradients of the bottoms of the fjords would 
be less than the gradients of the rock bottoms of the rivers, of which 
the fjords are continuations. For the gradient of a valley is always 
greater nearer its head than in its lower part, if the rocks are not greatly 
different in the two parts of the course. Further, the rivers of the 
coastal plain meander in wide valleys, now filled to a considerable 
depth with alluvial deposits. The width of these valleys compared to 
their depth, where known, shows that they were well advanced in the 
cycle of erosion, and that their gradients must have been compara- 
tively low. If this be true for the valleys of the coastal plain, to a 
much greater degree must it be true for the submerged parts of these 
valleys across the continental shelf. From what is known of the dimen- 
sions and contours of these fjords, a low gradient can be confidently 
predicted, unless the continental shelf has been tilted seaward. 
From the data given, the gradient of a few of these fjords can be esti- 
mated. The Bahaman valley has an average fall of twenty-four feet per 
mile for 350 miles; the Floridian, of twenty-five feet per mile for 400 
miles ; the Cazones, of seventy feet per mile for seventy-five miles, and the 
Altamahan, of nineteen feet per mile for 300 miles. These gradients are 
much greater than those of corresponding land valleys. The Grand Canon 
of the Colorado has an average fall of less than eight feet per mile for 
284 miles.* Judging from borings in the alluvium deposits of the Mis- 
sissippi at New Orleans and Memphis, the rock bottom of the river 
between these two points has a gradient of but little more than two 
feet per mile. Since it is not certain that these borings reached the 
deepest part of the valley at their respective localities, the gradient may 
be more than this ; it may, however, be less. But in spite of this element 
of doubt it is absolutely certain that the gradient is much less than that 
of the submerged valleys. The gradient of the fjords, considered in 
respect to the size of these valleys and compared with the low gradient 
of land valleys strongly favors the hypothesis that the subsidence was 
much greater remote from the present coasts, than it was along the 
shore. It cannot, therefore, be safely conciuded that because the 
mouths of these fjords are now submerged to depths from 10,000 to 
*DuTTon: United States Geological Survey, Monograph ILI., p. 240. 
