494 DIRECTION OF CHANNELS NEAR MICHIGAN BLUFF. 
these points there are more or less heavy masses of highly quartziferous gravel which appear to be 
entirely distinct in origin, and probably also in time, from the darker metamorphic gravels in ad- 
jacent channels in the same ridges ; and when we note the additional fact that the thickness of 
these guartziferous banks wherever they do occur at these localities is generally far greater than 
that of the dark-colored gravel which fills the other channels, and seems to underlie so extensively 
in thin and irregular sheets the volcanic capping, — we may be tempted to infer that such was 
actually the fact. But this would perhaps be going somewhat too far. For I have not been able 
to trace any well-defined channels here of this quartziferous gravel for any distance ; and further- 
more, I do not know its relative age, compared with that of the darker gravels, nor is it certain, 
indeed, that this relative age is the same at all the different localities. I suspect the quartziferous 
gravels to be generally older than the others, but I do not know. 
Where the large stream went to next from Michigan Bluff is also a question of much uncertainty, 
and it seems very probable that it may have followed, at least to some extent, different courses at 
different times. But there are many facts which lead me to think it probable that there was at 
least a period of time during which, after continuing its course a little to the west of south for a 
short distance farther, through Sage Hill, and passing out over the site of the modern cation in the 
vicinity of the present mouth of the North Fork of the Middle Fork, and meeting near this locality 
with one or more other streams, which probably then as now drained the central and southeastern 
portions of the present upper basin of the Middle Fork of the American, it curved to the west, and, 
rounding the spurs whence now come Mad and Ladies’ cafions, turned somewhat then to the north 
of west, and, crossing Volcano Cajon, entered the Forest Hill divide at Bath, and then, after fol- 
lowing a little way farther on this course, turned sharply again southwest and followed that course’ 
along the great channel which seems to exist in the Forest Hill divide, probably meeting also 
somewhere in this divide with the northern stream from Iowa and Wisconsin hills. 
One of the little facts which, so far as my knowledge of this country yet extends, seems to lend 
strong probability to the assumption that the drainage of the central portion of the upper basin of 
the Middle Fork once found its way to Bath, is the occurrence in the banks at that locality of 
numerous boulders of a peculiar variety of rock of the same class as the “ white lava” of Placer- 
ville, of which there are great quantities in place in the country about Long Caiion, but of which 
I saw none at all in any form in the country to the north of the Middle Fork of the Middle 
Fork. 
Furthermore, no stream could ever have passed from the hydraulic claims at Michigan Bluff 
directly through the ridge to Volcano Cajion and so across to Bath ; for the bed-rock in this ridge 
is everywhere far too high to admit of that. It must therefore have rounded the spurs below. 
But again, in the Forest Hill divide itself, it seems impossible (at all events, with the aid of only 
such facts as can now be gathered) to trace out any exact system for the streams which seem to 
have once united in this vicinity, and say, ‘‘ This is the exact course they followed,” at any particu- 
lar time. If the North and Middle Forks of the American River ever really united in this 
divide, as the probability seems very strong that once they did, there is plenty of further evi 
dence to lead to the inference that both their point of union and the detailed courses of the streams 
themselves have shifted about extensively during the long period of the slow accumulation of the 
auriferous gravel, aside from all the wanderings to which they may have been subject afterwards, 
during and after the volcanic period. 
The distribution of even the deepest accumulations of metamorphic gravel here among the 
branches of Brushy Caiion, at Yankee Jim’s, Todd’s Valley, the Dardanelles, and other localities, 
is such as no single system of streams, which other facts can render either plausible or admissible, 
is capable of accounting for. The evidences of change are everywhere. It is probable, indeed, 
that a large stream once flowed for a long period of time southwesterly along the great deep 
channel which seems to underlie the central portion of the ridge at Forest Hill, and it may be that 
this channel is continuous to-day beneath the ridge as far southwest as Peckham Hill. But again, 
there is evidence in the furrowing of the bed-rock at Yankee Jim’s, that two large streams once 
