62
Geological Survey of Finland, Bulletin 395
Tapani Mutanen
The lower contact is transgressive, climbing
the stratigraphy away from the feeder, which I
assume was in the western part of the intrusion
beneath the deep basin filled with thick olivine
cumulates (see map, Appendix 3). A gabbro
dyke projecting northwest from this basin may
be an extension of the feeder.
The original contact hornfelses of the intru-
sion are hardly distinguishable because of su-
perimposed retrograde (in relation to contact
metamorphic facies) regional metamorphism.
Hornfelsic sedimentary xenoliths have been
found in the lower part of the intrusion, and
relict contact metamorphic garnet occurs at the
top of the granophyre. The magma melted and
swallowed immediate roof rocks (high-alumi-
nous pelitic schists and arkosic quartzites) for
several hundreds of metres. Signs of incipient
melting are seen in the granitic clasts of an
older conglomerate immediately beneath the
intrusion in the western Kiviaapa area. As will
be discussed later, floor melting, rarely noted
in layered intrusions, was intense among the
salic floor rocks below the deep cumulate ba-
sin in the western part of the intrusion.
Shallow holes drilled recently (1995) in the
granophyre cap area, in the southeastern part
of the intrusion have intersected partially di-
gested acid volcanic rocks in which quartz
amygdales are often preserved (factually, as an
insoluble refractory residue) in a matrix with
the mineralogy and chemistry of granophyres
(SiO
2
54.35 – 65.78%, CaO 3.04 – 7.24%,
MgO 0.88 – 4.79%). Refractory roof rocks
(basalts, basaltic tuffs, various komatiites) oc-
cur as xenoliths, typically at or above the ma-
jor stratigraphic breaks and reversals.
There is no evidence of multiple magma
pulses at Koitelainen, nor does the petrological
explanation of the intrusion need them. As evi-
dence of fresh pulses I value only cases in
which magma is factually and unambiguously
seen to have intruded from beneath and to have
torn through cumulates. Considering the erod-
ing power of density currents sweeping along
the cumulate surface (see later) I attach little
significance to cumulate autoliths as evidence
of entry of fresh magma into the chamber. In
my opinion the intrusion formed essentially as
a single cast, before deposition of any signifi-
cant amount of cumulates. I consider the rever-
sals, often used as The Evidence for multiple
magma pulses, a problem to be solved, not a
solution to be satisfied with.
In analogy with the Akanvaara intrusion, the
parent magma (the magma that entered the in-
trusion chamber) was silica-saturated or just-
saturated, low-Ti tholeiitic in composition.
Fine-grained chilled microgabbros were
formed against the walls. Field evidence from
Koitelainen, Akanvaara and elsewhere sug-
gests that this lining did not last for long but,
loosened by the melting of the support rocks,
was peeled of and fragmented and left to
founder in magma (e.g., Hess, 1960; Morse,
1969a; Wadsworth, 1988; Wiebe, 1990). Be-
fore their true nature was revealed (Batashev
et al., 1976; Lavrov et al., 1976), such de-
tached, stoped chill slabs, often very extensive
(and, naturally with at least one sharp contact),
were earlier perceived as dykes.
New chilled rocks promptly formed in place
of the peeled-off margins. Stepwise chill re-
newal from continuously fractionating magma
may give a deceptive impression of successive
magma pulses (Mutanen, 1989b, Hoatson et
al., 1992). Later, we shall see how reversals in
the layered igneous sequences, which are gen-
erally (and too casually) ascribed to new mag-
ma pulses, might be explained by contamina-
tion.
In principle, a succession of temporary
chills would give a good measure of the evolu-
tion of magma. However, chilled rocks are
well known to be porphyritic, i.e., they contain
primocrysts (of olivine, pyroxene, plagioclase;
see e.g., Page, 1977; Hoover, 1989). Still
worse, chilled rocks are often mixed with con-
taminants (op. cit., and the following).
In the Koitelainen intrusion, autoliths of
chilled rocks are common in the Upper Zone
(UZ; analyses 3–4, Table 3). Most of them are