CASTLE CREEK. 263 



its jointing planes and surfaces with oxide of iron, but apparently free from 

 iron pyrites and metallic minerals. These quartz ledges must, however, 

 contain, at least, traces of gold, and furnish it to the placer gravels. Small 

 bright-green crystals of ripidolite in flat scales are inclosed in the mass of 

 the quartz; and green chlorite, full of red garnet crystals, incases the veins 

 when traversing the talcose and chloritic schists. Proceeding down the 

 valley the rocks gradually change in character, becoming less micaceous, 

 with a greater development of talcose and quartz schists, in places approach- 

 ing a clay- slate. 



Castle Creek, having made a most remarkable bend of eight miles to 

 the north, the greater portion of the distance through a narrow and deep 

 canon, resumes its former southeast course in a valley excavated among the 

 quartzites and clay-slates of the second division of the Archaean. An 

 attempt was made to explore this canon by myself and miners; but owing 

 to the prevalence of fallen timber and tangled thickets of small trees and 

 bushes filling the bottom of the gorge, we were unable to traverse it with 

 horses beyond a point about seven miles below where Custer's trail leaves 

 the creek. The exact line separating the two formations, the schists and 

 the slates, was not seen, but the dip was observed changing from west to 

 east, the quartz veins becoming less abundant in the rocks and free from 

 mica or chlorite, and more continuous and regular in formation. Thin 

 quartz veins followed the jointing planes of the slates as well as the strati- 

 fication of the formation. The quartz was locally ferruginous and cellular 

 from the decomposition of pyrites, and had more of a mineral-bearing ap- 

 pearance. It is evident from the change in character of the rocks that the 

 line of contact of the mica-schists and clay-slates must pass in a general 

 northwesterly direction from Spring Creek, near its union with Newton's 

 Fork, across the hills and the canon of Castle Creek until it is concealed 

 beneath the limestone ridge between Crook's Tower and Custer Peak. Cas- 

 tle Creek enters the canon in its north bend among the mica-schists, form- 

 ing a continuation north of the rocks of the French Creek district, and 

 emerges among clay-slates and quartzites identical with the rocks of the 

 lower valley of Spring Creek, and easily traced across the hills to that dis- 

 trict. Near the head of a small branch of Castle Creek, entering it from 



