RELATION OF THE IKON ORES 301 



ment appears to have been formed partly of loose fragments of siderite 

 or of silicious iron-range rock changed to oxides in their present posi- 

 tion, and partly of limonite deposited in a stalagmitic way from solu- 

 tions which probably encountered here a current of water charged with 

 oxygen. 



Professor Willmott considers that the cavities in which the ore was 

 deposited were formed by the folding of the rocks and solution while the 

 iron-range rocks still covered what is now the valley to the extent of 

 hundreds of feet above its present level. This does hot explain, how- 

 ever, the way in which the overlying rock was removed from the valley 

 and prevented from filling the two lake basins. 



Following this view, we might imagine all the iron to be removed in 

 solution, when not precipitated by meeting oxygenated currents, as at 

 the present ore body ; but the removal of the silica, which exists to the 

 amount of 5 or 10 per cent in even the purest parte of the siderite, is not 

 accounted for, and one would expect to find the rock basins filled with 

 finely granular silica, like certain bands of pure-white sand occurring in 

 the ore body as a residuum from the solution of the siderite. 



In accounting for the replacement of silica by iron ore in the Michigan 

 mines, Professor Van Hise* and other geologists have supposed that 

 alkaline waters have been the chief agent, and there is no doubt that 

 silica may be removed in that way, especially when the water is warm ; 

 but it is not easy to imagine such solutions at work after the valley con- 

 taining the two lakes here described was open to the sky, and the basins 

 could not have been completed before that time, or the fragments of the 

 overlying undissolved part of the iron range must have filled them when 

 the final collapse took place. 



There is no known source of alkaline solutions in the valley at pres- 

 ent, for the iron range, which includes most of the inclosing rocks, is 

 singularly poor in alkalies, and the immediately adjoining quartz-por- 

 phyry schists are themselves largely silicified and charged with car- 

 bonates. 



It is possible that the valley was cleaned out by the action of a stream 

 flowing through it at a remote time when the surrounding country had 

 not been cut down to its present level, but there is no evidence at hand 

 to prove this, and the. carving of the basins below the level of the outlet 

 of the valley could not be accounted for in this way. On the whole, it 

 seems more probable that the slow action of the weather and of the little 

 stream representing the local drainage cleared out the valley after sub- 



*U. S. Geol. Survey, Mori, xxviii, Marquette region, p. 403. 

 XLV— Bull. Geol. Soc. Am., Vol. 13, 1901 



