ORIGIN OF THE IRON ORES. 87 
The ores at and near the andesite contact are determined in their 
shape and distribution partly b}^ replacement of limestone, partly 
by peripheral tension fissures and other joints and faults, partly by 
faults later than the ore deposition, and finally by erosion. A calcu- 
lation of tlie radial shrinkage of the andesite mass in cooling makes 
it between 200 and 500 feet, depending on the thickness assigned to it. 
A part of this shrinkage is probably accounted for in the stretch 
fissures and faults in the andesite. A part of it has with equal 
probability furnished space for ore deposition at the periphery. 
At tlie sides of the andesite these spaces would remain open during 
and following their development, whereas on the slopes and on top 
gravity would tend to close them as fast as developed. The deposits 
therefore should have their maximum width where the contacts are 
vertical or nearly so and should thin out where the slopes become 
flatter. The conception of the authors as to the relations of the fis- 
sures to the andesite before erosion took place is illustrated in fig. 8. 
This figure does not take account of modifications of shape of the ore 
bodies due to replacement of the limestone or due to fissures normal 
to the andesite periphery, which are believed to be subordinate. 
Erosion has now cut down sufficiently far on the eastern and south- 
eastern sides of the several laccoliths to expose vertical or nearly ver- 
tical contacts. On the southwestern side of the Granite Mountain 
and Iron Mountain laccoliths the erosion plane has exposed a very 
gently dipping contact plane between andesite and limestone. The 
wide extent of the deposits at these places, expecially at the south- 
western side of Iron Mountain, is largely due to the fact that the 
erosion plane is so nearly parallel to the plane of contact. It does not 
indicate that the deposits are necessarily wider when measured in a 
direction normal to the contact than they are elsewhere in the dis- 
trict; indeed, if fig. 9 represents approximately the shape of the 
openings which determined the ore deposition, the deposits here may 
well have been thinner than elsewhere. Under these conditions ver- 
tical faulting has greater effect on the surface distribution. 
It follows from the above considerations that the maximum depth of 
the ore deposits near, the contacts may not be greater than the depth 
to which the and esite-limes tone contact extends, and this is deter- 
mined by the thickness of the andesite laccolith and whether it breaks 
across the limestone layers or has tilted them up in such a manner that 
the contact is approximately parallel to the bedding. In view of the 
fact that the limestone is tilted steeply where erosion has exposed the 
sides of the laccolith, it may be assumed that the limestone may be in 
contact with the andesite as far down as the laccolith goes. Therefore 
the thickness of the laccolith becomes the determining factor. Ero- 
sion has allowed no direct means of measuring this thickness in this 
district, for the bottom is nowhere exposed. A comparison of the hor- 
