36 
MUSEUM BULLETIN NO. 30. 
as a whole, although the major number of the specimens examined 
carry bytownite, still enough are found to make clear the 
tendency of the feldspar to ’become more and more sodic. Out 
of thirty-seven sections of olivine gabbro examined, six, or 16 
per cent, contained feldspar whose composition varied from 
A.b 30 An 70 to Ab 40 An 60 ; and in thirty-three sections of augite 
gabbro, thirteen, or 40 per cent, carried feldspar between Ab so 
An 70 and Ab e5 An 35 . That these sodic feldspars are really products 
of a later crystallization than the more basic is shown by the 
facts : that they are found in the outer bands of zoned crystals 
where these occur ; that the augite gabbro, already proved to be 
a differentiate of the olivine gabbro and slightly later than it in 
crystallizing, contains a larger proportion of types carrying the 
more sodic feldspars; that the augite gabbro carries feldspars 
much more sodic than the olivine gabbro; and that the granite, 
proved a still later differentiate, is characterized by feldspars 
more sodic still. 
The possible changes in the composition of the pyroxene 
could not be determined satisfactorily with the microscope. As 
for the olivine, most of the crystals show partial resorption. 
The mineralogical changes accompanying the appearance of 
quartz in the rocks have been described, and, like those of the 
feldspars and the olivine, follow exactly the lines indicated by the 
investigators of the geophysical laboratory. These facts indicate 
that as crystallization proceeded some agency continuously re- 
moved the crystals from the liquid magma. Preceding sections 
have shown that the agent must have been gravitation. 
Special Textures and Structures. 
The structure of the anorthosites may be specially cited as 
most easily explicable by the hypothesis already given. As 
previously mentioned, they are found as a rule in small isolated 
masses, a foot or so in diameter. These masses are usually 
coarsely crystalline, and contain very little admixture of the 
other rock minerals. It is difficult to account for such occur- 
rences, except by the theory that, as feldspar crystals separated 
from the magma, they were segregated from the olivine and 
