266 TRANSACTIONS OF THE [MAY 18, 
production of minerals is not confined so closely to a limited 
area, still the well-developed crystals are limited to a small part 
of the contact zone,and there can be no doubt that their 
formation was largely dependent upon the presence of heated 
solutions and gases. The abundance of apatite is in harmony 
with this conclusion, this mineral being so often a product of 
pneumatolytic action. 
In both cases described, the large size and perfection of the 
crystals may result in part from the influence of mineralizers, 
as suggested by Williams* for pegmatites, and in part from a 
slow growth. The influence of mineralizers can hardly be ques- 
tioned, while a rather slow crystallization is also probable. At 
the time of intrusion the surrounding rocks would be quickly 
heated, and the normal contact zone would develop as a result. 
But heated solutions would continue to circulate during the 
entire period of slow cooling, and along the lines of ready trans- 
mission the limestone would gradually give place to new minerals. 
Such conditions would, judging from experimental data, be 
favorable to the growth of large crystals. 
Before passing from these clearly defined cases of the forma- 
tion of important mineral masses by contact metamorphism to 
more doubtful and obscure occurrences, brief mention may be 
made of previously described localities.+ The well-known 
minerals of Natural Bridge are, for the most part, on the con- 
tact between limestone and an intrusive rock of variable com- 
position, which, in the absence of careful investigation, has been 
tentatively called gabbro. The minerals of these contacts are 
pyroxene, orthoclase, mica, scapolite, wollastonite, titanite, 
zircon and calcite. This is an association quite similar to that 
of the Rossie localities, though lacking apatite and tourmaline, 
and having instead wollastonite and zircon. The contact nature 
of the masses is perfectly clear and the same concentration of 
mineral development at certain points is well shown. 
In the localities thus far described, the association of 
species is such as to suggest at the outset their formation by 
contact metamorphism; and in the foregoing pages the effort 
has been made to show that this is actually the case. The data 
derived from these clearly-defined instances afford a basis for 
the consideration of occurrences where the phenomena ac- 
tually shown render a direct determination of origin impos- 
sible. The analogy is sometimes so complete that conclusions 
*G. H. Williams, General Relations of the Granite Rocks in the Middle Atlantic 
Piedmont Plateau. XY. Ann. Rept. U.S. G.S., p. 684. 
+ C. H. Smyth, Jr., The Crystalline Limestones and Associated Rocks of the N. W. 
Adirondack Region. Bull. Geol. Soc. Am. VI., p. 279. 
