2(18 'THE PENOKEE IRON-BE ARING SERIES, 



within tlie usual siliceous groundmass may he cut by ramifying veinlets of 

 silica. Again, areas are seen in which curving lines of iron oxide have been 

 developed at or near their edges. These Hues may or may not comjilete 

 a loop. If the outline of the ttriginal carbonate area is more or less irreg- 

 iilar'aud angular, the iron oxide curve cuts off the irregularities. In other 

 cases, again, several such complete or })artial iron oxide lines, concentric 

 with one another, have been developed. Subsequent tti or alternate or simul- 

 taneous with this process silica appears within the concretion, filling the 

 spaces between the rings of iron oxide, the iron carbonate being previously 

 removed. At other times the space occupied by the iron carbonate is left 

 vacant, or else only partially filled, whence arise the frequent geodic and 

 other cavities found in this phase of rock. Finally, no iron carbonate 

 remains, its place being taken by the concentric rings of iron oxide and the 

 niiimtely crystalline to amorphous silica. Between the unaltered iron car- 

 l)onate areas and the perfectly formed concretions there are at times, often in 

 a single section, every possible gradation. If the iron oxide developed from 

 the carbonate was very plentiful but little room was left for the entering 

 silica, and these areas appear now nearly or quite opaque with oxide, but 

 generally the iron oxide has been developed in the partial or complete 

 ovals referred to, and the substitution of silica for the remaining carbonate 

 left these lying within a siliceous background. 



Quite similar has evidently been the process by which the pseudo- 

 fragmental areas, such as shown in PI. xxiii, Figs. 1 and 2, have been 

 ])i'oduced. In such cases the silica of the background has plainly also been 

 rearranged, so that it resembles and merges into the silica of the concre- 

 tionary areas, thus giving the vague, irregular outlines which make these 

 forms so closely resemble fragments. More rarely the fragmental areas 

 are tabular in form, or else show a subordinate parallel lamination which 

 appears to be lines of original deposition. The lamiiiEe of these areas 

 frequently abruptly terminate at the exteriors of the apparent fragments. 

 A number of such areas sometimes occur in a single section, now lying 

 within the siliceous gi-oundinass in such positions as to have the lamina^ of 

 the different areas make all sorts of angles with each other. In such cases 



