526 THES VOCE 
The olivine here shows no sign of idiomorphism, but this is often 
the case with labradorite. Laths (parallel to oro) of the latter here 
and there (Fig. 24a) protrude into the olivine, and the straight-edged 
boundary which we often observe (Fig. 23) between the labradorite 
and the olivine is not caused by the crystal-limit of the olivine, but 
by that of the plagioclase. The feldspar must thus here have 
commenced crystallizing at an earlier stage than the olivine, which, 
in the same manner as the diallage in Figures 24a and 246, and as 
the hypersthene in Figure 14, chiefly forms an intervening mass 
between relatively large labradorite individuals. 
In the oliine-rich olivine hyperites (hyperitic-structured 
olivine gabbros) with about 25-30 per cent olivine, 60 per cent 
AY 
0G 
K E S= =f 
’ ‘Ah Z 
\ Ai i 
Fics. 24a and 24b.—Drawings (20:1) 
Anorthosite from Ekersund, Norway, containing ca. 90 per cent labradorite 
(Ab;An,), ca. 7 per cent olivine (dotted in Fig. 24), and a little diallage (dark shading 
in Fig. 24, not seen in the photograph). The straight lines in Figure 23 represent 
the idiomorphic contours of the labradorite against the olivine. 
labradorite, 10-20 per cent diallage, and a little magnetite, etc., 
the olivine chiefly appears in synneutic individuals with very good 
idiomorphism against the diallage and partial idiomorphism also 
against the plagioclase (see, for instance, Fig. 33). Considering 
only the silicate minerals, we find consequently that first a good 
deal of olivine solidified, then the labradorite, and at a later 
stage the diallage also commenced crystallizing. 
In the olivine-poor olivine hyperites, with only 5-10 per cent 
olivine, we find, on the other hand, in the relation between the 
olivine and the labradorite quite a different structural phenome- 
