170 
THE GEAJTD GAEOX DISTEIOT. 
no definite or assignable relation to them in any respect whatever. One 
very marked ravine comes from the summit overlooking Little De Motte 
Park, and runs along the middle of the promontory which ends at Point 
Sublime and terminates at the bi’ink. No ravine enters the Shinumo at its 
head, hut two enter it at the sides. There is no ravine entering the head of 
the western branch of the Hindoo, but a rather deep one enters the head of 
the eastern branch. A conception of the real state of the case may, perhaps, 
be gained by recalling the fact that these amphitheaters were originally 
narrow cuts produced by corrasion, and subsequently widened by the 
weathering of the edges of the strata thus exposed. The walls recede by 
waste away from the axis of the cut, and sooner or later a ravine here and 
there is tapped laterally by the receding wall wasting backwards across its 
course. Thus we often find an old ravine suddenly cut off on the brink of 
an abyss, and the continuation of the same ravine is seen upon the other 
side of the amphitheater. Quite probably, at some very early stage in the 
excavation of the Grand Canon and its lateral gorges, the amphitheaters 
were merely the lower courses of the stream beds which constitute the 
Kaibab drainage system. But these streams dried up and the enlargement 
of the chasms went on in a measure independently of the distribution of 
the surface ravines. Here and there a ravine shows that it has maintained 
the old relation and enters the amphitheater at its head. Such ravines are 
almost always the largest ones, and they notch the wall of the amphitheater 
to a depth of 400 to 600 feet. They may in such cases possibly be regarded 
as the main channels of the lateral gorges. But usually the surface chan- 
nels end anywhere upon the brink in seeming caprice as to choice of locality. 
If we could restore the mass which has been removed in the Grand Canon 
we might very probably find them gathering together into a series of ordi- 
nary tributaries, constituting a limited number of district drainage trees, 
the trunks of which are the existing channels at the lower ends or openings 
of the great amphitheaters. 
From the Milk Spring we may also make a journey southward to the 
end of the great promontory which forms the eastern wall of the Hindoo 
Amphitheater. The scene here presented is of the same general character 
as thatfi’om Point Sublime. It is not quite so satisfactory, however, because 
