Chap. XXXVITl. ] sandub state. 
1009 
picked up by a magnet. It also gives a fairly strong reaction for 
manganese, showing that this dement is present in appreciable 
amount. It is not, of course, jjossible to be sure that this manganese is 
an original constituent and was not intioduced by percolating solutions ; 
especially when there is such an abundance of manganese-ore close 
at hand, i ut the rock seems to be fresh. It is only a portion 
of the rock in this exposure that has the characters given above. The 
remainder has suffered a partial alteration of a very interesting 
character. The altered rock, which is illustrated in the accompanying 
figure (77), appears variegated in patches of red, white, and black. The 
black part is almost invariably separated from the white by a zone of 
red colour, as indicated in the figure. A careful study of the rock in situ 
b = blank ; 
r = red ; 
w = white. 
Fig. 77. —Piece of altered magnetite-slate. 
has suggested that the probable explanation of the phenomenon is that 
under the influence of solutions saturating the rock the iron and 
manganese have started to segregate. The manganese has segregated 
faster and retreated into the thin black zones marked ' b ' in the figure. 
[Occasionally the very centre of the black band is occupied by a thin 
veinlet of psilomelane, marking the final stage of concentration.] The 
iron has retreated less rapidly than the manganese, forming red zones (r) 
on either side of the manganese bands. In so doing the iron, which 
was partly in the protoxide form in the fresh rock (as magnetite), 
suffered peroxidation with the formation of the red colour. The 
white, part represents the rock bleached of its manganese and a 
part of its iron ; for it has not lost all its iron and gives a slightly 
reddish streak, due to the fact that the portion of the iron that has 
not retreated into the red zones has been peroxidized n the same 
way as that which has retreated. In one place the beginning of the 
change is seen in the appearance of a small red area in the midst of the 
otherwise fresh rock. Doubtless this patch would have continued 
