321 
3. There are no apophyses passing from one sheet into 
the next. 
The junction of the naujaite with the lujavrite is excep- 
tional. In some places there may be found a slow and gradual 
transition, but more usually this junction is represented by a 
characteristic breccia zone which will be considered more fully 
below. 
From these contact relations it is inferred that all the 
sheets are of approximately the same age, and that most of 
them have probably been in a semi-fluid or viscous state at 
the same time. Furthermore, the tolerably horizontal disposi- 
tion of the sheets seems to indicate that gravity was the main 
factor which controlled the form and the distribution of the in- 
dividual rock bodies. 
The most simple assumption, therefore, which at once 
lends itself as a possible explanation of the sheeted structure, 
is that we are dealing with the results of differentiation in 
place; or speaking more precisely, that which was originally a 
more or less homogeneous magma, has, while yet in a semi- 
fluid condition, separated itself under gravity into horizontal 
sheets of different compositions. Upon closer examination, 
however, the question proves to be more complicated, but be- 
fore it can be discussed more fully it will be necessary first 
to consider some other facts relating to the structure of the 
batholite. 
The concave shape of the sheets. — In the above state- 
ment, the batholitic sheets have been spoken of as approxi- 
mately horizontal. Their actual form more exactly can be char- 
acterized as that of flat saucers piled up upon each other (Fig. 
29). This is seen very clearly in the natural sections, especially 
in those on the two sides of Tunugdliarfik Fjord (see p. 68, Fig. 
ве ата Bl VI, Fig. I): 
The behaviour of the stratification at the boundaries of 
the batholite is variable. At some places the sheets are at 
XXXVIII. 21 
