524 Ee OG 
as in the crystallization of deep-seated igneous rocks, we find a 
different case. Forellengestemm or troktolite, for instance, consists 
of very basic plagioclase (anorthite-bytownite) and olivine. Some 
anorthosites show predominant plagioclase (bytownite, labradorite) 
and in addition some olivine, and some olivine gabbros consist 
chiefly of basic plagioclase and olivine with only a small amount of 
monoclinic or orthorhombic pyroxene. On remelting several of 
these rocks at atmospheric pressure, more or less melilite results,. 
and in addition frequently some spinel, and occasionally also other 
minerals. At the pressure of one atmosphere, consequently, quite 
a different mineral combination results from the magmas here 
mentioned than the combination occurring in the deep-seated 
rocks. We shall in a following chapter discuss the cause.of this. 
Here we shall only fix our attention on the fact that the melting 
of intermediate mixtures of CaAl,Si,O, and Mg,SiO, at the pressure 
“of one atmosphere does not give us the required information of 
the individualization boundary between plagioclase (anorthite) and 
olivine, which takes place at high pressure. In the question in 
hand, therefore, we must use the analytical method. 
Petrographical experience shows that the olivine in most of the 
igneous rocks crystallized at a very early stage. Because of this 
fact, many petrographers have the conception that olivine takes an 
exceptional position with regard to the sequence of crystallization 
and that this mineral always crystallizes as No. I of the silicate 
minerals. This is, however, a misconception. The reason why 
the olivine in numerous cases commences crystallizing earlier than 
the other silicates is that the individualization boundary between 
olivine and plagioclase or pyroxene, or between olivine and plagio- 
clase plus olivine, lies at the point of relatively little olivine, and 
that most rocks contain more olivine than the individualization 
boundary mentioned. But if less olivine is present the solidification 
commences with the crystallization of one of the other minerals. 
Returning to the rocks which consist only or nearly only of olivine 
t As early as in my first mineral-synthetic work (“Studies on Slags,” 1885) 
I opposed this misconception, which inter alia also has been expressed in petro- 
graphical works of later years. But when a misconception of an authoritative 
character has been impressed on one’s consciousness, it may require a decade of years 
to uproot it. 
