Origin of Pegmatite. — Crosby and Fuller. 159 
In many cases the inclusions, especially of schist, have an 
exceedingly frayed-oiit and skeleton-like appearance, suggest- 
ing a ditferential solvent action and shredding rather than 
mere mechanical breaking or cracking. The wall itself is 
often similarly affected, as well as more or less distinctly im- 
pregnated with tourmaline and other minerals of the first 
crystallization. This is perhaps the most significant of all 
the contact phenomena coming under our notice. 
We have not observed an}"- clear examples of the centric or 
spheroidal structure described by L. von Buch and others, and 
so strongly insisted upon by Brogger ; nor has the rare fluidal 
structure cited by Brogger* come under our notice. But 
we have noted one fact which very plainly suggests motion 
in the niass during the crystallization of the pegmatite. In 
a small vein in the Winnepesaukee gneiss north of Honey 
Brook, in Acworth, the tourmajines, which extend about two- 
thirds of the way across the fissure, are for a distance of 
many feet tilted uniformly from the perpendicular to the wall 
about thirty degrees. This tilting is real and cannot be ex- 
plained by any irregularity of the walls. 
External, forms and formal relation.'^ to inclosing rocks. — 
With pegmatite, as with pretty much all subterranean forma- 
tions, form is less distinctive and significant than structure 
or composition; and the morphologic argument is rather 
evenl}'' divided between the rival theories — aqueous and igne- 
ous. In general, the occurrences are dike-like in form- but 
they may with equal propriety be described as vein-like. 
No doubt the spaces or cavities occupied by the pegmatites 
are chiefly mechanical in origin — spaces of discission; and 
they nnist, therefore, be conditioned by the structure of the 
inclosing rocks and the nature of the disrupting force. The 
more regular masses may be described as rectangular (Figure 
4), wall-like (vertical sheets), lenticular (Fig. 5, PI. VIII), or 
tabular (horizontal sheets) ; but a very large proportion of the 
occurrences are too irregular to come under either of these 
terms (Fig. 6, PI. IX). They very commonly conform closel}'^ 
with the structure planes of the schistose rocks (interbedded 
veins) ; but they also intersect llie bedding planes or lamina- 
tion at all angles; and may in general be assumed to follow 
lines of least resistance to the disrupting force. 
*Canadian Record of Science, 6, 64 and G5. 
