OKIGIN IN IGNEOUS ROCKS. 
83 
The interior veins, I, II, and III, in fig. 12, which at the present 
time have no connection with each other, but are each separate and 
distinct, w^ere at the time of their formation part of the corundum 
concentrated along the border of the peridotite. Some of these veins 
would soon be worked out, wliile otliers might be explored for a Inm- 
dred feet or more without any apj)arent change in their width. 
This explanation will account for jdl the variations in the occur- 
rence of the corundum in the ditferent veins. 
The corundum crystallizing out from the molten peridotite, which 
Avould be a very basic nuignia. Would be conceutrated toward the 
•^^ ' 11 - V « -.. = '' ^ - „ = 'I I ihV II I : ',1 1, iiiiMii I 
iPIiii 
fill i;' 
«ts;i!ilil;!lilii!lll! 
III!' 
mmk 
.1 fi 
! I'i i:l^V-'^>; 
l!r ll '"■•■ 
i I! |'''V 
I I ''-^'•'' 
i iiiiiiiiiii;jir> '' ='- ^ 
ii!iili;i!l!!;lil'l--:i- 
■---'^--^-^^>-^-^v^|lll'll III 
il • vv^ 
;ii III! 
5 llifi=^: 
'iai'iiii!i:i!iiiii!ii-..'/ 
'%":r'H!«tlif'' 
"'" " - " =11 ^'i r iii|''i"l 
:>■;! Il 
■ .-V;4 I! f 
Fig. 12. — Ideal vertical cross section of mass of peridotite soon after its intrusion into 
gneiss. 
margins; and though in many cases there Avoukl be a sharp separation 
(at the time of formation) between the corundum and the peridotite, 
in others there would be more or less corundum, in the form of masses 
and particles, that would extend beyond the main masses into the peri- 
dotite and thus give the appearance of a somewhat gradual passage 
from corundum to the pure ])eridotite, as illustrated in hg. Kl 
The varying pressure, temperature, and other physical conditions 
during and after intrusion would ail'ect the crystallizing and the 
separating out of the mineral fi-oui the molten mass, and this will 
explain the great variation obscrN'ed in the corundum fouud in the 
same mass of peridotite. Thus, at Corundum Ilill, North Carolina, 
