DSSTRIBUTION AND $TRATIGRAPHIC
FO‘SIHON Q? LATE PRECAMBRMN
DIABA$E DSKVES IN PARTS OF
NOR‘E‘HE‘SRN MECHEGAN

 

$352333 5:0:- ”in Degree of M. S.
IEiECHIGAN STATE UNEVERLSETY
Warren ‘W. Wood
1962

 

LIBRARY

Michigan State
University L

 

 

 

 

 

 

 

 

 

 

f

”at! cm mamas"!

wmm or GECLOG?
an unsmc, MICHIGAN

 

 

DISTRIBUTION AND STRATIGRAPHIC POSITION
0F LATE PRECAMBRIAN DIABASE DIKES
IN PARTS OF NORTHERN MICHIGAN

BY

Warren W. Wood

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of GeOIOgy
1962

(T 2

7 ') ‘fi 1/ ,f'
4‘ urn/”(3"

// /,..‘< /

ACKNOWLEDGEMENTS
The author wishes to express his sincere appreciation
to Dr. Justin Zinn who initially suggested the problem,
and gave so freely of his time during its undertaking.
Deep appreciation is also extended to Dr. James Trow
and Dr. William Hinze for their time, suggestions and con-

structive criticisms of the manuscript.

11

CONTENTS

Page
Abstract......................................... 1

Introduction.....................................

I Purpose and Scope...........................

Location and Physiography...................

Field Work and Methods......................

O\U'|Ull\)f0

Previous Work...............................
Geology of the Area......................... 7
Distribution of Fresh Dikes...................... 10
Description of Samples........................... 15
Petrography................................. 15’
Location and Field Relationship............. 18

Possible Sources of Error in Model Conver-

Sionoooooooooooooooooooooooooooooooooooooooo 27

Compositional Similarity of Keweenawan Flows and
the Late Bikes.I.0.0000000000000000000I0000...... 29

Structural Analysis.............................. 35
Reverse Polarity................................. 45
Economic Importance.............................. 52
Conclusion....................................... 54

Rafarences CitedOOOOOOOOOOOOOOOOOOOOOOOOOOOOCOOOO 55

111

INDEX OF TABLES

Page
Table 1. Chemical Analysis of Lighthouse Point

Dike...OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 21

2. Comparison of Dikes (Thin Section Ana-
ly318).................C...’............ 33

3. Chemical Comparison of Keweenawan Lava
Flows with the Late Dikes............... 34

iv

ILLUSTRATIONS

Page

Figure 1. Location and General Geology of Study

AreaOOO..0....0...OOOOQOOOOOOOOOOOOOOOOO

2. Location of Samples Studied............. 17

3. Photomicrograph of Typical Diabase
Structure...0.000000000000000000.0.0....19

4. Photomicrograph of Fine Grained Diabase
Sthtureooococo-coco.oooocooooooooooooo 19

5. Photomicrograph of Altered Diabase
DikeSOOOOOOOO0..OOOOOOOOOOOOOOOOOOOOOOOO 25

6. Photomicrograph of Ophitic Keweenawan
Lava FlOWOOOOOCOO...OOOOOOOOOOOOOOOOOOOO 25

7. Photomicrograph of Skeletal Magnetite... 28
8. Photomicrograph of Skeletal Magnetite... 28

9. Illustration of Strike of Fresh Dikes
in Parts of Northern Michigan........... 37

10. Illustration of Late Diabase Dikes Rela-
tive to the Lake Superior Basin Axis.... 44

11. Photomicrograph of Possible Powder Mag-
netitBOOOO0.0.000.000.00.000...000...... 51

Column 1. Generalized Geologic Column for Parts
Of Northern MiChiganOOCCOOOOCOOOCOCOO... 9

Plate 1. Distribution of Late Precambrian Diabase
Dikes in Northern Michigan
FaCIngPageOOOOOOOOOOOOOOOOOOOOOOOOOOOOO15

ABSTRACT

The late Precambrian diabase dikes of northern Michi-
gan outcrop in Gogebic, Iron, Baraga, Marquette and Dickin-
son counties. They are strikingly similar to one another
in field characteristics and composition. Most of the
dikes are magnetically reversed to the present geomagnetic
dipole and intersect with a definite trend all other rocks
in the area. The distribution of these dikes is not fully
known but they appear to lie within at least a 50 mile wide
belt which appears to parallel the Lake Superior Basin axis.

The dikes probably intersect the post-Animikie granite
and are overlain unconformably by a probable Lower Cambrian
sandstone. They could therefore be either (1) a post
granite pre-Keweenawan intrusive (2) genetically related
to the Keweenawan sequence or (3) post-Keweenawan - Pre-
cambrian. Structural evidence indicates that the dikes
occupy positions probably created by the same stresses
that formed the Lake Superior Basin. Chemical similarity
between the dikes and the Keweenawan lavas is shown. Re-
verse polarization is investigated as a possible method
of correlation and is rejected.

It is concluded that the dikes are Keweenawan in age.

INTRODUCTION
Purpose and Sc0pe

An extensive amount of field mapping and exploration
work, undertaken throughout the older Precambrian areas in
the western portions of the Northern Peninsula of Michigan,
has shown the presence of great numbers of basic dikes cut-
ting these older rocks. These dikes vary in direction, size
and other characteristics. Some of them appear to be impor-
tant in iron ore occurrances. Most of these dikes clearly
show the effects of metamorphism in that they are now urali-
tized or chloritized and in some bodies they have apparently
altered to clay minerals. Frequently, however, a dike rock
is found which in contrast to most of the others, appears
surprisingly fresh and unaltered. Where carefully mapped,
these dikes cut all lower and middle Precambrian rocks.
Other characteristics of these fresh dikes are noteworthy:
most are reversely polarized, that is, magnetic readings
over them are negative compared to those over adjacent rocks.
These diabase dikes have been found in most of the iron bear-
ing districts of the Northern Peninsula either as surface
outcrops or intercepted in drill cores. Aeromagnetic surveys
have also located some of the larger dikes and these show
dominant trends cutting indiscriminantly through all other
rocks.

Many geologists have speculated as to the age of these

dikes and usually stated that they were probably Keweenawan,

but little detail work has been done on them. iost previous
investigation has been fragmentary and usually associated with
some economic inquiry. Consequently, little has been written
about these dikes as a system.

The study of these late Precambrian dikes was undertaken
in order to determine their distribution and stratigraphic
age. It is held by the writer that a system of dikes, such
as exhibited in this area, should be part of a larger orogeny
and not a separate entity. The process of fracture and in-
trusion of dikes on such a large scale should be related to
some larger orogenic activity and consequently should bear
structural, chemical and/or mineralogical relationships to
this activity. 4

Under the circumstances, comparison and contrast of the

chemical composition of the late Precambrian dikes with post-

Animikie and Keweenawan intrusions in conjunction with a
structural analysis was considered the best method of strati-
graphic correlation. This entailed sampling for chemical
composition and areal mapping to determine dip, strike and
stratigraphic relationship.
Location and Physiography

This study is basically confined to five counties in
Northern Michigan: Marquette, Baraga, Gogebic, Dickinson and
Iron (Figure #1). The dike system probably extends through-
out the Lake Superior Province but time did not permit such
an extensive investigation.

The topography of the area depends generally upon the

FIGURE

 

‘

so by

 

 

 

catafioooi

.\ \QN \‘ , I- / {G 3.3.. as. 22::

 

 

 

 

 

 

 

 

 

<\o<m<o\ ., L/ . ////// // ,/,,/ s
/+//// zoo<ZOkzo , . 5.5600 032.03
Wfi/kfi kfifl / /.,/,./._,,, // // // oEou. one 53:338.

9%
4 )
e 4, meme
3:... o..o.1 m. m m__

240.10.: ZKUIkmoz mo
Q43. Ichmxm UEOJOMO

 

 

resistance of the exposed rocks. In contrast to a series
of parallel ranges common to the Keweenaw Peninsula, the
five counties exhibit numerous irregular hills, lakes and
swamps. In Marquette, Felch and Menominee regions a defi-
nite series of east-west trending ridges are prominent.
Parallel valleys are formed from less resistant formations.
Field work and Methods

Exposures of Middle Precambrian rocks in Northern
Michigan are spotty and generally lack continuous exposure
due to heavy cover of glacial drift and vegetation. This
condition in conjunction with the large area of this in—
vestigation required extensive motor travel. Fortunately,
Northern Michigan has an abundance of good Federal, state
and secondary roads which makes the area readily accessible
by automobile.

Field work for this problem was conducted in the
summer of 1960. The dikes under investigation were recog-
nized by their distinctive brown weathering, their reversed
polarization (as observed by dip needle) and their unaltered
appearance. Each dike location was examined for dip, strike,
post emplacement movement, effects upon country rock (struc-
tural and mineralogical) and magnetic anomaly. A represen-
tative sample was taken of each occurrence investigated.
All laboratory work was conducted at Michigan State Univer-
sity and consisted of thin section study and qualitative

chemical analysis.

Previous Work

As early as 1381 Rominger recognized the similarity of
the fresh dikes of Northern Michigan with the "Copper bearing
Formations" (Keweenawan). Irving and Van Hise (1892) felt
that the dikes of the Gogebic district were intimately con-
nected to the Keweenawan lava flows. They state that the
dikes were very probably the pipes which fed the lava flows.
Van Hise and Bayley (1897) stated they thought the fresh
dikes of the Marquette district were intruded during Kewee-
nawan time (p. 218). Clements et alu(1899 p. 189) called
attention to the similarity of the dikes of the Marquette,
Gogebic and Crystal Falls areas with each other and with the
Keweenawan lava flows. He also warned that chemical similar-
ity was a hazardous method of correlation. Van Hise and
Leith (1911) mention that fresh dikes are found in all areas
surrounding Lake Superior.

Royce (in Newhouse 1942) studied these dikes in several
locations in connection with enrichment of the iron ore.
Balsely, James and Wier (1949) commented on the negative mag-
netic anomaly these dikes produced. Graham (1953) studied
the reverse polarization of several dikes in Baraga county.

Many of these workers assigned these dikes to the Kewee-
nawan simply on the basis of lack of metamorphism as compared
to the other Precambrian rocks in the area. While this method
may be satisfactory it does not allow for the possibility of
a post-Animikie, pre-Keweenawan period of intrusion which,
on the basis of stratigraphic relationship, the dikes could

occupy.

7

Geology of the Area

The Precambrian geology of Norhtern Michigan is ex-
tremely complex and is treated here only as it pertains to
the late dikes. The various districts have experienced
similar events of sedimentation and diastr0phism and only a
brief description is given.

Most of the area has at the base of the middle Pre-
cambrian a quartzite and dolomite succession overlying un-
conformably the lower Precambrian. This is followed by a
series of quartzite, slate and iron formation, which in turn
is overlain by a succession of quartzites, iron formation,
volcanics and slates. This last sequence is apparently re-
peated in some areas. All of these formations are intruded
by dikes and sills which are now metadiabase, and folded
contemporaneously with intrusions of granite (James et a1,
1961). A profound unconformity followed as indicated by
the depositidn of lower Keweenawan beds upon the folded and
plained older rocks as shown in the Gogebic district. Radio-
metric methods of dating, indicated about 300 million
years between the ages of the post-Animikie granite 1.4 X 109
and the Duluth Gabbro 1.1 X 109 (probable Keweenawan) (James
1958). The dikes of this study intruded the Animikie series
and granite, and in the Gogebic range some of the lower Kewee-
nawan flows.

In Michigan, the Keweenawan series (upper Precambrian)
is represented by basic lava flows interbedded with felsic

conglomerates. The series is generally divided into 3 units

on the basis of lithology: lower, middle and upper. The
lower and upper units are primarily sedimentary with some
basic flows in the lower and more feldspathic flows in the
upper. The middle unit is composed mainly of basic lava
flows which extend 15,000 feet in some areas. The extru-
sions and sediments are confined mainly to Keweenaw, Hough-
ton and Ontonagon counties and form the Keweenaw Peninsula.
Unconformably above the Keweenawan, lies a probable
Cambrian sandstone (Hamblin 1958) upon which, unconformab-
ly, till and outwash of the Pleistocene epoch is deposited.
A generalized geologic column is given for the middle Pre-
cambrian. A more detailed description of the upper Pre-

cambrian is also included (Column 1).

GENERALIZED GEOLOGIC COLUMN FOR
PARTS OF NORTHERN MICHIGAN

 

 

 

 

 

 

 

 

 

 

Precambrian

Cenozoic Quartenary Pleistocene outwash and till
Paleozoic Cambrian Jacobsville sandstone
Freda sandstone
Upper Nonsuch shale
—Copper Harbor group
Upper Middle Eagle River group
Precambrian Keweenawan Ash Bed group
series Central Mine group
r Lowe; Bohemian Range group
Granite (Intrusive contact)
Metadiabase (Intrusive contact)
Middle Upper
Precambrian Animikie
series Middle
—Lower
Granite, Syenite, Peridotite, Gneiss,
Lower

Metasediments

 

DISTRIBUTION

The system of late Precambrian dikes under study
(Plate #1), apparently represented by two varieties, prob-
ably exists throughout the Lake Superior province. In
Michigan, middle and lower Precambrian exposure areas-afford
the only opportunity for open examination due to the over-
lap of Paleozoic sediments from the southeast and Kewee-
nawan-Cambrian sediments from the northwest, wherein dikes
can be found.

Both reversely and normally polarized fresh dikes are
observed in Northern Michigan. Whether the two types truly
represent two different times, different swarms, or different
composition, is not known; it is not intended to imply simi-
larity of intrusion for all reversely polarized rocks. This
feature is used for comparison purposes like any physical
property and may not have a time significance. Both types
of dikes intersect the post-Animikie pre-Keweenawan granite,
but no normally polarized dikes have been observed truncated
by the Jacobsville sandstone. The reversely polarized dikes
appear to be much more abundant, however, this may be due in
part to their reverse polarization which makes them immediate-
ly recognizable.

In the Gogebic district of Michigan, late Precambrian
dikes are extremely numerous, occurring in all of the mines
of the areas, occasionally 3 or 4 in a mine. The dikes found
here are generally altered by the fluids that concentrated

the iron ore. Fresh exposures, which are similar to late

10

11

dikes in other areas, are found only away from the ore.

It is probable that the original magnetism in the altered
areas, regardless of its orientation, was destroyed by the
ore concentrating fluids. There are no published aero-
magnetic maps of the region in which to check for reversals
of polarization, and the writer was unable to find any out-
crops to check this feature; however, dip needle traverses,
run in 1926 by the Michigan Department of Conservation,
indicate some negative reading over diabase outcrops. These
are interpreted by the writer to be an expression of the
same dike system as observed in the other areas of Northern
Michigan.

Graham (1953) indicated 25 probable reversely polarized
dikes in a 18 X 18 mile square in Baraga county. An aero-
magnetic survey (Balsley, James and Wier 1949) on parts of
Houghton, Baraga and Iron counties, covering additional
area to Graham's report, shows a prominent belt of nega-
tive magnetic anomalies between Township 47 North and Town-
ship 49 North. When correlated with the geology, these
negative anomalies are found to be associated with fresh
diabase dikes. The dikes appear to start abruptly at Town-
ship 49 North and to decrease southward from Township 46
North (Plate 1).

In Marquette county there are at least 10 reversely
polarized dikes and probably several times that many, but
the aeromagnetic surveys which have been flown by the United

States Geological Survey and would have disclosed these re-

12

versely polarized dikes, have not been released. The
area of intrusion could be predicted for Marquette county
on the basis of the intrusions in Baraga county and in-
vestigation by the author. The largest number of dikes
would be expected to lie between Township 46 North and
Township 49 North and strike E—W with little variance.
All of the samples collected by the writer in Marquette
county (Figure #2) are within this belt; in addition, a
fresh olivine diabase dike of unknown magnetism and age
occurrs on Green Island in Lake Michigamme (Van Hise and
Bayley 1897).

Dikes south of Township 46 North might be eXpected
to strike N 70 E, if they follow the basic fault pattern
of the Marquette synclinorium (Empire, Volunteer and
Palmer faults). Three reversely polarized dikes have
been observed striking N 70 E in the southern area. The
dike at the abandoned Isabella mine (Sample #7). is in a
fault plane parallel to the main faults of the area.

This dike exhibits minor slickensides which are attributed
to post emplacement readjustment. Buzas (1960) found 2
negative anomalies (R 24-25 W., T 47 N and R 23-27 W.,

T 43 N) (Plate #1), which he attributed to reversely
polarized dikes. These dikes parallel probable faults
(Allen 1914). While Buzas did not intimate that the dikes
were intruded into fault zones, it is possible that the
strong negative anomaly produced by the dikes masked the
magnetic break generally exhibited by a fault. If the

13

strike of the dike in B 23-27 W., T 43-44 N, is projected
into Dickinson county it coincides with a series of faults
north of the Felch trough in R 28-30 N., T 42 N.
In the Mineral Hills area of Iron county there are
a few fresh dikes, one of which is reversely polarized
and fault controlled (James and Dutton 1951). Gair and
Wier (1956) found one reversely polarized fresh dike in the
Kiernan Quadrangle (T 43 N., R 31 N.). Bayley (1959) found
only a few fresh stringers a few inches to 3 feet wide in
the Lake Mary Quadrangle, south of the above area (T 42 N.,
R 31 W.). None of these fresh intrusions were reversely
polarized. An aeromagnetic survey (Balsely, James and Wier
1949) disclosed the presence of several reversely polarized
dikes in T 45 N., R 34-35 N., and also one in T 42 N., R 35 N.,
in addition to the previously mentioned dike at Mineral Hills.
An aeromagnetic survey of Dickinson county (Wier, Balse-
ly and Pratt 1953) disclosed only two magnetic anomalies
which could be interpreted as reversely polarized dikes.
In central Dickinson county, extremely detailed mapping dis-
closed one large reversely polarized dike (mentioned above)
and several small fresh normally polarized dikes (James
at. al., 1961). There is no mention by Bayley (1904) of
fresh dikes or intrusives in the Menominee district, nor
did the aeromagnetic survey of Dickinson county disclose
any reversely polarized dikes in that area.
It can be seen that the dikes were intruded in a belt
at least 50 miles wide, with the main belt of intrusion

14

25 miles wide (Plate #1). Most of the area of intrusion
is occupied by folded Animikie sediments. To what extent
these folded sediments may have controlled intrusion is
unknown but it appears that many dikes away from the main
intrusion belt are controlled by preexisting faults and
should be considered in this light.

It would be expected that the belt of dikes in Baraga
county is an extension of those in the Gogebic district.
Whether an aeromagnetic survey could confirm this under a
cover of Cambrian sediments and Keweenawan traps is doubt-
ful. This belt undoubtedly extends through Marquette county
and eastward into Lake Superior. Final evidence rests on
the release of the aeromagnetic data from the United States

Government.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

    
 
 

     

 

 

 

 

 

 

 

 

 

 

 

m Boundary of Aeromognetic Surveys

9 IRON MOU NTAIN

20° ‘ 7 h 7 s9° as ‘l
ONTONAGON I 'L I
! 57/ WW I
L i A l .9
_ ,, _. __z.__j , ,, _ _ c __ (’20
I T i (T/ I l 6‘
i i K I i ' i '9/
J I l O ’ i I 0
I i l i 4 L i 6)
i s , — ~ - I— /,i c _ i ,
Q t— s ‘ I ,2 , . i \
ONTONNGON I ;I l) MAiRQUETTE :49 “K
. e ".2 I si- I
I 3 , , T
i l i A J - =48 ' MARQUETTE
— " W“ ‘ W — —“"‘“ ’ f O N ' " Ari/i“ i " ““1" ' .,..,,-__ ,,_ 1 W“ \\ \\\\\ ’_/\_—
i Six/i , x‘vj I ‘ I ‘
i I i I r l : 4 M i 2
* l .K ‘ \
T“ l T ].___ W W ix— ,1 4 /L,_ ., __ +7 _ 322:7- .,
I j Y . 4 , \
E I I F I E { I f
42 j 41 4o 39 38 37 1 36 I\ 35/” 34 g 33 i 32 F 31 3o 29 I 28 ? 27 26 :25 24 23 /
r i 7 ’ _ ”—0—— .- ‘. __i—‘k EA L\ i /i V I T M” ii fl ._._c,_i,.g,, ’ i \\\K‘ 1‘
f i A E ' f : ’ i /
C { K ; , 1 , 1
I i r\ / /7 I ; [\7'/'/'/"/'/‘l'r//7777317/"/‘/777/7/‘/A; 45 \
,,i ,i -1 .3. ;, __.__v_ i r f i \ i / \ z /
, if V i‘ *7 W" “W —' WW ~ +W ' - - —' — " i
‘ ‘ 4 r ! 1 ‘ x ii ‘i , “V 31 w I ‘
\j\\ I J Ix R; o) N. o Igc KlNSOfN 44
‘ i j \ i / I 3 \ 2 A: /
i‘ i 1 1 1 / 1 [\ 1 1 \
\ i 3 Y
DISTRIBUTION OF LATE PRECAMBRIAN N L\‘ ' ; i: —j%"' j j j \ I ,’ \43 I #7;
DIABASE DIKES IN NORTHERN MICHIGAN \ . I I\ 2 “5E? .\ i 1; \l/I/1/111/////14_'l/1 // ’
scale ~/ i j i K I 4 l
O 10 :0 i ”i 42 46"
must "3 i I '—
’ ' ' T x: T
—— Reversely Polarized Dikes ~ i 6 4| I
o °°°°°° Normal and Unknown Polarized Dikes [ £3 I”
fin

 

 

 

 

 

 

ESCRIPTION OF SAMPLES
Petrography

The use of the integrating stage with calculation Of
modal percentages was considered adequate for the purpose
of this study, allowing both a petrographic and semiquan-
titative chemical analysis. Over 300 equalgranular crystals
were counted in a typical mode, thereby indicating a standard
deviation Of about one percent of that represented by the
slide (Niggli 1954). Moorhouse (1959) believes that when
optical properties are known with sufficient accuracy, the
volume percentage method is more accurate than conventio-
nal chemical analysis. Correlation of optical properties
with chemical composition was accomplished by the use Of
the findings of Hess (1949), Larsen and Berman (1934),
Moorhouse (1959) and Johannsen (1939), the latter work
being used in converting species composition to oxides.
Conventional petrographic techniques were used in Obtaining
the Optical data.

When altered minerals were encountered in a modal tra-
verse, the original mineral was recorded if its identifi-
cation was positive. In instances where there was in-
sufficient evidence as to the character of the original
mineral, the altered mineral was included in a separate
category but not included in the conversion to oxides, as
the original composition was the primary interest. In most
instances the degree of alteration was minor, thereby mini-

mizing the danger of selective alteration changing the per-

15

16

centages of any particular mineral.

The fresh reversely polarized dikes (Sample Locations
Figure #2) have a megascopic texture varying from near
glass at the contact edges to a courseness of grain size
in the center that is dependent upon the width of the dike.
This clearly indicated that cooling proceeded from the con-
tact edge. The major mineral constituents in these dikes
are plagoclase, pyroxene, olivine and magnetite. A light
rusty brown color is characteristic of the weathered sur-
face. On fresh unweathered surfaces the rock is a gray-
green to black.

MicroscOpically, the fresh holocrystalline diabase
dike (Figure #3) consists of 50 to 60% idiomorphic, carlsbad
and albite twinned, labradorite lathe (Ab 44-46), 30 to 35%
hypidiomorphic clinopyroxene (species pigeonite), 0 to 5%
Olivine (Fa 10-30), and 5 to 10% titaniferous magnetite.
The latter three minerals are interstitually deposited.

The minerals are generally fresh and unaltered; only the
olivine suffers minor alterations on its periphery and in
irregular cracks throughout the crystals. The average
grain size of the samples taken is 0.4mm, if the long di-

mension (O.5mm-2.5mm) Of the labradorite can be neglected.

17

FIGURE 2

 

 

 

 

F C
35303: e m .
O
mImIIV m
N
3:...
ON 0. O
I J
325

202304 oEEom.

whhmaomflz

L

__‘

I'—
_.__.____J

N HEDGE

20m:

i
I
i
i
I

<o<m<m

a.

$26.5

—' 1

 

 

18

Location and Field Relationship
Sample 1

Location: Lot 6, Can. 5, Township of Mckim, District of
Sudbury, Ontario, Canada. The dikes Of this type are
vertical and trend NW in the Sudbury Basin. They are the
last intrusives in the area and only rarely are they cut
by faults. They are observed intersecting the Murray
Granite (Thomson 1956) and the Killarny Granite (Cooke
1946). The sample was taken from a dike that was exposed
intermittently over % mile. The width, about 75 feet,
did not vary much in that distance. These dikes in the
Sudbury Basin have a positive magnetic inclination and
only when they pass through a magnetic formation are they
reversed (Zurbrigg 1961). This sample was collected for

the comparison of late intrusives in other Precambrian

areas.
Volume Percentage gg’Mineggl Species
Pigeonite ------- 15.75 Olivine ------- 3.91
Magnetite ------- 6.70 Labradorite--- 73.63
Mineral Species Conyerted 32 Oxides
$102 A1203 Fe203 FeO MgO CaO Nazo
50.14 20.21 4.89 6.45 4.23 8.44 4.12

Sample 2
Location: NEfi, NW% of sec. 24, T. 48 N., R. 25 W, Corner
Of Pine St. and East Michigan St., Marquette, Michigan.

This outcrop is a western extension of Sample 5 as traced

Figure 3

Sample 5, X-Nicols 20 power. Typical fresh
diabase as exhibited by most examined dikes.
Lath shaped Carlsbad and Albite twinned feldv
spar, hypidiomorphic clinopyroxene (pi eonite),
olivine (minor in this photomicrograph? and
magnetite.

Figure 4

Sample 4, 50 power plane light. Fine grained
border Of Sample 5. This illustrates border
chilling and feldspar development.

19

Figure 3

4

 

r f

19"}? '.

s ‘\ ‘5
‘ -_ V
. . \“\\ . _

20

by dip needle. The 25 foot wide outcrop, dipping 60°
north, was sampled in the center. No contact edge was
visible to ascertain the metamorphic effects upon the

surrounding Mona schist.

Volume ngcentagg p§_Min§pal Specigs
Pigeonite ------- 34.89 Olivine (Fa 20) ------- 1.17

Magnetite ------- 4.19 Labradorite (Ab 45) ---59.74

Mineral Species Converted pg Oxidgg
$102 A1203 Fe203 FeO MgO CaO Na20
50.42 17.62 3.49 11.48 3.78 7.93 3.25

Sample 3 and 4

Location: NE%, NWi, of sec. 24, T. 48 N., R. 25 W, at
Lighthouse Paint in the city of Marquette, Michigan.

These samples were originally collected to determine
if there is a change in composition with respect to the
contact edge. The samples were collected 1 and 8 inches
respectively from the contact edge. The crystals, upon
examination, were too fine grained for consistent Optical
identification (Figure #4). The plagoclase however, had
albite extinction angles comparable to the sample Ob-
tained in the middle Of the dike (Sample #5).

Lane (1911) describes chemical analyses made from
samples at Lighthouse Point. The contact, .616mm, 4.115mm
and 7.6mm are reproduced in Table #1. A comparison Of

chemical versus thin section analysis is included. The

21

Table 1

$102 .élée 1% 11-25 1180 $39
A1203 17.85 17.55 18.00 17.47 17.40
Fe203 3.13 2.51 2.21 2.66 5.08
FeO 10.36 12.69 12.42 12.93 10.86
MgO 7.16 5.65- 6.35 6.88 5.69
CaO 8.47 10.75 11.45 10.27 7.71
Na20 2.04 2.21 1.96 1.91 3.21
K20 .60 .65 .66 .59

H20 4.47 2.07 ---- ----

T102 -—-- ---- ---- ----

P205 .14 .17 .16 .16

CO2 ---— —--- ----- ----

S .09 .18 .08 .11

SO3 --—- ---------------

MnO .26 .19 .18 .15

Cl .07 .05 .02 .09

Dike at Lighthouse Point (Lane 1911)
(1) Contact
(2) 0.616 mm from margin
(3) 4.115 mm H "
(4) 7.6 mm " "
(5) Thin Section Analysis 4 meters from margin (Sample#5)

22

results are so nearly the same that the use of thin sec-
tion analysis for comparative purposes is justified.
Lack of mineralogical variation along strike is afforded

by comparing Sample #2 and #5.

Sample 5
Location: NE%, NW% of sec. 24, T. 48 N., R. 25 W, at
Lighthouse Point in the city of Marquette, Michigan.

The 25 foot wide outcrop is well exposed at this
location where it intersects the MOna schist and other
dikes cutting this schist. This is the only observed dike
that deviates from vertical by dipping 72° north.

Volume ngcentagg p§_Min§pal Species
Pigeonite ------- 30.87 Olivine (Fa 1o) ----- 3.43

Magnetite ------- 6.59 Labradorite (Ab 45) - 59.12

Minepal Spgcigg Cpnpgrtgd 32,0xiges

$102 A1203 Fe203 FeO MgO CaO Na203
49.09 17.40 5.08 10.86 5.69 7.71 3.21
Sample 6

Location: SW&, sec. 21, T. 48 N., R. 27 W, 1.1 miles north
of Deer Lake in Marquette county.

The vertical east west dike intersects the Kitchi
schist at this location. Slight baking Of contact was Ob-
served. Thehigh percentage of alteration in this sample
is probably due to surface weathering. The hand specimen

shows a definite increase in weathering toward the surface.

23

The high percentage of magnetite may be due to alteration
Of iron rich olivine but most of the magnetite is primary.

Vplume Percentage pf,Mineral Species

Pigeonite ------- 29.96 Olivine (Fa ?) ------- .44
Magnetite ------- 9.84 Labradorite (Ab 45) --52.37
Alteration ---- 7.39

Mineppl Spgcies Convgrtgd pg Oxideg

47.45 15.68 7.82 12.51 4.29 7.49 3.07
Sample 7

Location: SE&, SW% of sec. 29, T. 47 N., R. 26 W, at the
abandoned Isabella Iron Mine.

A vertical dike, striking north of east, intersects
the Negaunee Iron formation at this location. The sample
appeared to have slickensides on it but no microscopic

evidence Of brecciation was found.

Volumg Perpgntagg p; Minepal Spgcigs
Pigeonite ------- 34.29 Olivine (Fa 25) ------ 6.73

Magnetite ....... 7.54 Labradorite (Ab 45) -- 51.41

Mineral Spgcies Cpnverted pp Oxides

$102 A1203 11.203 FeO MgO an ha203

47.81 15.11 5.78 13.76 6.65 6.88 2.87

24

Sample 7 (a)
This sample was obtained from the Isabella mine dump
and thought to be the altered equivalent Of Sample 7. The
rock is completely altered with only relic fabric and some

plagoclase remaining (Figure #5).

Sample 8
Location: SE%, NEi, NE& of sec. 7, T. 48 N., R. 27 W.
north of Ishpeming, Michigan.
The steeply dipping (by dip needle) east-west striking
dike probably intersects the Michigamme slates, although

the surrounding rocks were not exposed for identifica-

tion.
Volume Percentage pf Mineral Species
Pigeonite ------- 30.14 Olivine (Fa 30) ------- 5.10
Magnetite ------- 6.97 Labradorite (Ab 43) --- 57.67
Mineral Species Converted pp Oxides
$102 A1203 F8203 FeO Mgo Cao Na203

47.93 17.19 5.43 13.77 4.11 7.77 3.00

Sample 9
Location: NE%, NW%, NW% of sec. 18, T. 49 N., R. 33 W,
near the town Of Alberta, Michigan.
The vertical, south Of east trending dike, 25 meters
wide, intersects the Michigamme formation at this location.
The sample was taken 4 meters from a contact edge. The

development of large plagoclase crystals gave the dike an

Figure 5

Sample 7a, 40 power plane light. An example
of a dike altered by iron ore concentrating
fluids. Only relic fabric and some magnetite

is visible.

Figure 6

Sample 10, 40 power X—Nicols. Typical ophitic
texture displayed by the more basic Keweenawan

lava flows.

 

26
"Oatmeal" look, however, the dike was not porphyritic.

Volume Percentage 2; Mineral Species
Pigeonite ------- 19.60 Olivine (Fe 25) ------- 2.36

Magnetite ------- 5.38 Labradorite (Ab 46) ~-- 72.65

Mineral Species Converteg _p Oxides

50.22 21.20 4.05 7.99 3.02 5.78 4.15

The variation of mineral percentage, compared to
that of other dikes examined, can be explained without
postulating a change in the composition of the dike.

The statistical validity of this thin section was in-
validated by the development of large feldspar crystals.

Consequently, this dike is not included in the average.

Sample 10
Location: SE&, SWé, NE% of sec. 1, T. 54 N., R. 34 W,
a Keweenawan lava flow near the geology building at the
Michigan School of Mines at Houghton, Michigan.

This section exhibits the ophitic texture which is
typical of the less feldspathic Isle Royal flows (Figure
#6). No species percentage was taken on this sample, as
the statistical condition of granular equality was not
kept. The minerals are augite, labradorite, olivine and
magnetite. This section has not suffered from the mineral-

ization as have other flows.

27

Sample 11

This sample was taken from an unknown flow in a
quary outside Houghton, Michigan. The sample is com-
pletely altered; not even the texture is discernible.

Possible Sources of Error
in Modal Conversion

There are several possible sources of error in con-
verting the modal analysis to oxides, even when the op-
tical properties are known. The Optical data of olivine
was derived from the center of the crystal which is prob-
ablt richer in magnesium than the periphery of the crys-
tals due to the sequence of crystallization. The data
derived from examination of the feldspars may not repre-
sent the true composition due to some zoning on large
crystals. This would tend to increase the total $102
and Na20 and decrease CaO and A1203 from the true com-
position. Thirdly, the ilmenite included in the magne-
tite was computed as FeO and Fe203 in conversion to ox-
ides. The extent of ilmenite can be seen in Figures #7
and 8 which show the skeletal magnetite crystals after
removal of the less resistant ilmenite. The excess of
computed iron in the magnetite overcompensates for the
lack of computed iron in olivine, probably giving a high-
er total iron than actually exists, and the titanium con-

tent is not computed.

Figure 7
Sample 6, 40 power plane light. Magnetite

with ilmenite weathered out, leaving skele-
tal structure.

Figure 8

Sample 6, 40 power plane light. Same as
photomicrograph 5.

28

Figure 7

‘ O

 

 

COMPOSITIONAL SIMILARITY OF KEWEENAWAN FLOWS
AND THE LATE DIKES

Several late normally polarized dikes have been ob-
served intersecting the post-Animikie - pre-Keweenawan
granite in northern Michigan. There is no known instance
of reversely polarized dikes intersecting this granite,
however, one such dike was observed (sec. 29, T.42N.,
R.28W.) a few hundred yards from the late granite, and had
suffered no thermal metamorphism. None of the examined
dikes indicated thermal metamorphism that James (1955)
has shown to be associated with the late granite. All
of the late dikes in northern Michigan are therefore in-
ferred to be younger than the post-Animikie - pre-Kewee-
nawan granite.

As to stratigraphic position, these dikes could be
either (1) a post granite-pre-Keweenawan age (2) gene-
tically related to the Keweenawan volcanic series, and
therefore of Keweenawan age, or (3) post-Keweenawan -
pre-Jacobsville in age. A chemical comparison is in-
cluded to help resolve this situation.

Clements et al., (1899) mention the similarity of the
fresh diabase dikes in different sections of northern
Michigan with one another. By observing these dikes in
the field it is easy to see how this conclusion was reached.
The dikes weather to a distinctive rusty brown that appears
to be diagnostic of these dikes. They occupy the same
stratigraphic position in all the areas and their strike is

29

30

generally constant for a given area. These and other
characteristics give the impression of similarity of intru-
sion among the areas.

Table #2 compares the composition of the samples
collected, all of which are extremely similar. Analysis
I and Y are taken respectively from a normal and a reverse-
ly polarized dike located in central Dickinson county
(James et al., 1961). They are left in the volume per-
centage figures due to a lack of Optical information on
the minerals. However, if the augite is pigeonite, as it
is in the examined dikes, the composition would be simi-
lar. Sample #9. from Alberta, Michigan, is statistically
unacceptable due to the development of large feldspar
crystals and is not included in the dike average compo-
sition, as it was included in the study for comparison of
late Precambrian dikes in other areas.

Table #3 gives a chemical analysis from a dike in
the Gogebic district (Irving and Van Hise 1892) and is
compared with an average of the analyses of the Lighthouse
Point dike given previously in Table #1. The Gogebic dike
is apparently more basic, however, the similarity between
the two dikes is striking. The similarity of composition
of the dikes of the Gogebic district with those of the
Marquette and Iron River districts is important. If a
similar magma source can be postulated for both of these
locations, thereby inferring a similar time of intrusion,
the dikes can be restricted to a limited time interval.

31

In the Gogebic district the olivine diabase dikes inter-
sect the Lower Keweenawan sequence (Irving and Van Hise
1892) and are therefore younger than Lower Keweenawan in
age.

Clements et al., (1899), while acknowledging the
similarity of these dikes, warned of correlation by chemi-
cal composition over extended distances. Comparison of the
analyses (Table #3) may not establish a true correlation;
they are intended only to show the possibility of the same
or a similar magma source.

Similarity of composition between the fresh dikes and
the Keweenawan lava flows first led Rominger (1881) to the
conclusion of equivilance of the two. This similarity is
illustrated in Table #3 which compares several flows with
the analyses from the Gogebic dikes; the average of ana-
lyses from the Lighthouse Point dike and the dike averages
as determined by thin section analysis. This comparison,
allowing for small variations caused by different analy-
zing techniques, illustrates similarity of percentages of
all important oxides. The difference between the indi-
vidual flows is greater than the difference between the
dikes and the flows.

These dikes, if they are Keweenawan in age, may not
have intruded at the same time in all areas. If the Lake
Superior Basin was sinking throughout the deposition of
the Keweenawan sequence (Hotchkiss 1923) fractures would

form progressively away from the basin axis as it subsided,

32

consequently, dikes further away would be younger. A
dike system which represents a certain arbitrary unit of
time, may actually have been intruded throughout a period
of many thousands of years.

If Butler and Burbank's (1929) conclusions, that the
flows become more feldspathic higher in the series, is cor-
rect, then, additional evidence of a Keweenawan age of the
dikes might be obtained if the dikes were more feldspathic
progressively away from the basin axis. Evidence for this
change is at the present time inconclusive.

As these dikes intersect the younger granite they
are obviously younger than this intrusive. If the dikes
are considered to be genetically connected to the late
granite, it is difficult to account for their distribu-
tion, which is unrelated to the outcrop area or zones of
metamorphism caused by the granite. Additionally, it is
not probable that this acidic magma (Table #3) of limited
extent, would have associated with it the large quantities
of basic magma now incorporated in the dikes. The com-
parison between the Keweenawan lava flows and the late
dikes offers nothing which would prevent the conclusion
that the flows and the dikes could be from the same magma
source. It, however, does not prove their equivalence.

In addition, the intrusion pattern of the dikes, covered
in detail in the section on structure, indicates a Kewee-
nawan stress pattern and therefore presumably a Keweenawan

age.

33

TABLE 2

A
l 2 5 6 7 8 9 262386

5102 50.14 50.42 49.09 47.45 47.81 47.93 50.22 48.54
.41203 20.21 17.62 17.40 15.68 15.11 17.19 21.20 16.60
2.203 4.89 3.49 5.08 7.82 5.78 5.43 4.05 5.52
F60 6.45 11.48 10.85 12.51 13.76 13.77 7.99 12.47
MgO 4.23 3.87 5.69 4.29 6.65 4.11 3.02 4.90
CaO 8.44 7.93 7.71 7.49 6.88 7.77 6.78 7.55
N820 4.12 3.25 3.21 3.07 2.87 5.00 4.15 3.08

Analysis X Analysis Y
Plagoclase 53.7 (an 54) Plagoclase 52.8 (an 54)
Augite 41.0 Augite 41.8
Magnetite 5.3 Magnetite 3.1

(l) Sudbury Ontario, Canada. Thin section analysis

(2) Pine St. Marquette Michigan. Thin section analysis
(5) Lighthouse point. Thin section analysis

(6) Deer Lake Marquette County. Thin section analysis
(7) Isabella mine. Thin section analysis

(8. North of Ishpeming. Thin seition analysis

(9) Alberta dike. Thin section analysis.

Analysis X Normally polarized dike north of Felch Mich.

Analysis Y Reversely polarized dike north of Felch Mich.

34

TABLE 3

.131. .12. m. .127 a. .14., 41:4. 71%.
A1203 15.60 17.71 16.60 16.34 16.82 12.63 13.98 15.46
Fe203 3.69 2.63 5.52 4.73 3.09 3.84 3.59 1.92
F80 8.41 12.10 12.47 6.13 7.91 10.30 9.87 2.46
MgO 8.11 6.51 4.90 6.43 6.65 5.23 6.70 .31
Geo 9.99 10.24 7.55 9.68 9.87 6.55 9.38 2.26
Na20 2.05 2.03 3.08 2.68 2.63 3.53 2.59 4.10
K20 .23 .62 .55 .48 1.90 .69 2.20
H201, 2.49 2.98 1.48 2.82 1.80 .42
CO2 .38 .19 .08 .26
T102 .82 1.47 1.53 2.50 2.19 .28
P205 .13 .16 .18 .18 .22 .32 .13
S .12 .Ol .02

Cu .009 .01

MnO .17 .20 .16 .16

(1)

Gogebic diabase dikes (Lane 1911)

(2)
(3)
(4)

Average of analyses of Lighthouse Point dike (Lane 1911)
Average composition of late dikes (thin section analyses)
Kearsarge flow group (average of 7 analyses) (Broderick
1935)
(5) Greenstone flow group (average of 12 analyses)
(Broderick 1935)
(6)
<7)

(8)

Bed #65 Eagle River, Keweenaw Point, Mich. (Lane 1911)
Daly's average for plateau basalt

Post4Animikie, pre-Keweenawan granite (James at al., 1961)

STRUCTURAL ANALYSIS

In attempting to date the fresh diabase dikes of the
‘western part of the Northern Peninsula of Michigan, these
facts have been noted: these dikes are truncated by the
Jacobsville sandstone of probable Cambrian age; they inter-
sect the folded Michigamme slate and probably the post-
Animikie granite. A structural analysis, to further de-
fine their age will be attempted. Such an analysis will
present several hypotheses since it involves comparing the
probable stress patterns of the Animikie and Keweenawan
and seeing which fits the dike intrusion pattern more
accurately.

The intrusion pattern should be explained in terms of
stress and forces responsible for the observed relation-
ships. Fractures bear a definite geometric relation to
the state of stress at failure. If the relation between
stress and fracture is known, then the probable fracture
system resulting from a given stress distribution can be
predicted (Hefner 1951, Sanford 1959). Conversely, given
a systematic fracture system it may be possible to de-
termine the distribution of stress that produced it. The
dikes may not occupy fracture positions but it is assumed
that zones of weakness would occupy the same geometric
relationship with the stress axes.

The dikes in eastern Marquette county strike east-

west with a variance of 50- 100 north or south. In western

Marquette county and easteranaraga county the dikes are

35

36

consistently a few degrees north of east as are the re-
‘versely polarized dikes in Iron county. A geologic map
interpreted from aeromagnetic data (Balsley, James and Wier
1949) indicates that this change continues through western
Baraga county where the strike approximates 10°- 15° north
of east. In the Gogebic district, dikes thought to be of
the same group, when rotated to the vertical, strike north-
east.

The dips of the observed dikes are approximately
vertical with the exception of Lighthouse Point, Pine St.
dike which dips 62°— 72° north. Aeromagnetic data from
parts of Houghton, Baraga and Iron counties (Balsley, James
and Wier 1949) indicates that the reversely polarized dikes
are nearly vertical in these areas. In parts of Marquette
county, magnetic interpretation by Buzas (1960) indicates
that the reversely polarized dikes were vertical. In the
Kiernan Quadrangle, Gier and Wier (1956) indicate that a
reversely polarized dike was steeply dipping. In the north-
ern part of the Iron River district, James and Dutton (1951)
found a fresh unaltered late Precambrian reversely polarized
dike that dipped 60°- 80° south, over a distance of a mile
and a half.

In the examined samples, the olivine diabase dikes are
vertical or nearly so, and the country rock shows no appar-
ent strike-slip or vertical displacement. There is only
one observed instant of shattering which might be inter-

preted as the result of shearing. From this evidence it is

38

inferred that the intrusion pattern of the olivine diabase
dikes was apparently controlled by horizontal tension. It
would appear that the dikes intruded in a wedging manner
such as described by Anderson (1942): the magma actually
propagating the displacement into which it flowed. The
Isabella dike with its N. 70° E. strike is probably fault
controlled (Royce in Newhouse 1942) as is the reversely
polarized dike in the Mineral Hills area (Balsley, James
and Wier 1949). The strike of the dikes located by Buzas
(1960) indicates that they probably follow a fault pattern
(Plate #1).

A case could be made for the proposition that the
strike of the olivine diabase dikes in the Marquette dis-
trict resulted from release of the thrusts from the north
or northeast that are believed to have folded the Marquette
Synclinorium at the end of Animikie time. Such thrusts
would give the synclinorium the observed, nearly east-
west strike. Release zones of weakness could then form
parallel to the axis of the Marquette structure in an east-
west direction. This hypothesis however, would not explain
the trend in areas other than the Marquette district.

A brief review of the major Keweenawan structural fea-
tures is given here so as to facilitate and understanding
of the relationship the intrusion pattern of the olivine
diabase dikes may have with the stress patterns of the Kewee—
nawan epoch. The most conspicuous structural feature

of the Keweenawan epoch is the Lake Superior Basin. This

 

' .‘Mv.

39

linear, asymmetrical structure has a strike of N. 600 E.
‘west of Keweenaw Point, extending in this direction into
southern Minnesota. East of Keweenaw Point the basin prob-
ably strikes close to SE, extending to the east end of
Lake Superior (Butler and Burbank 1929) (Figure #10). The
Keweenawan succession may extend as far southwest as Kansas
(Thiel 1956).

The Keweenaw fault parallels the main axis of the Lake
Superior Basin. The fault is of a reverse nature, with
the north block overthrusting to the south with varying
dips (17°- 60°). The length of the fault is in excess of
110 miles (Butler and Burbank 1929). This fault (which may
have formed during the intrusion of the Duluth Gabbro
(Hotchkiss 1923), at which time basin was uplifted and sub-
sequently collapsed) may have a throw of more than 3 miles.

The Lake Superior Basin was probably subsiding during
the deposition of the Keweenawan sequence. This is in-
ferred from the observed progressive flattening of the dip
in a stratigraphically up direction, and from the direc-
tion of the flow of the lavas (Hotchkiss 1923). This sub-
siding linear basin, which was probably formed as the result
of magmatic extrusion and subsequent subsidence (Hotchkiss
1923), would have certain stress patterns associated with
it, and these patterns would control intrusion during this
subsiding period.

Experimental data indicate that in macroscopically

homogeneous and isotrOpio rocks, tension fractures develop

40

parallel to the intermediate principal stress axis and are
perpendicular to the direction of the greatest stress axis.
“Using an idealized concept of a subsiding linear basin as
a model, it is possible to predict with some reservations,
the strike of the fracture pattern likely to have been
formed in the Lake Superior region during Keweenawan time.
The least stress axis would everywhere be perpendicular to
the axis of the basin. The angle of dip of the fracture
formed under these conditions would be controlled by the
dip of the plane of the intermediate and least stress axes.
Linear zones of weakness, if not failure, resulting from
tension would be expected to form parallel to the axis of
the linear element. Extension fractures or zones of weak-
ness, perpendicular to the basin axis, would also be ex-
pected. The system of dikes exhibited in northern Michigan
is probably subject to conditions which would cause de-
viation from the ideal. However, the system is fairly
consistent over a large area.

The olivine diabase dikes in Marquette county exhibit
the same strike as the theoretical tensional fractures
proposed for the area. This is also true in western Bara-
ga county where the change in strike, north of east, of the
dikes corresponds to the probable change in strike of the
basin axis at that point. The strike of the reversely
polarized dikes in Iron county corresponds to that in
Baraga county. In the Goulas River area of Ontario, Cana-

da, a pair of olivine diabase dikes, intersecting all other

41

IPrecambrian formations, strike N. 45° N. and N. 45° E.
(Moore 1946), corresponding to the theoretical fracture
pattern of the Lake Superior Basin. They are parallel and
perpendicular. respectively, to the probable basin axis.
In the Michipicoten area of Ontario, a late Precambrian
olivine diabase dike strikes N. 35° N. (Moore 1946). This
strike would approximate the probable strike of the basin
axis in this area. However, it is not known whether the
Goulas River or the Michipicoten dikes are of the same
chemical composition or if they exhibit the reversed polar-
ization of the dikes in northern Michigan. They are des-
cribed petrographically only, as olivine diabase. Van Hise
and Leith (1911) state that freSh diabase dikes are para-
llel and perpendicular to the shore line of Minnesota.

Late Precambrian olivine diabase dikes are mentioned
in the literature of the Gogebic Iron district of Michigan.
The Gagebic range has been tilted so that the Huronian
sediments are dipping 60° to 70° north. The dikes which
intersect the Huronian and some of the lower Keweenawan
flows dip south at angles of 20° to 30°. If the Huronian
sediments were rotated back to a horizontal position, the
dikes would be vertical. The corresponding strike upon
rotation would be northeast - southwest (Irving and Van
Hise 1892); this dip and strike would fit into the pro-
posed fracture pattern of a subsiding linear basin by para-
lleling the basin axis (Figure #10).

Hotchkiss (1923) believes that the Gogebic area was

42

tilted southward 20° to 30° when the first Keweenawan sedi-
:ments and flows were deposited. This area then gradually
tipped back toward the north as more sediments and flows
'were deposited; about Middle Keweenawan, the area was hori-
zontal again. It may have been at this point that the Go-
gebic dikes were intruded. That Middle Keweenawan was a time
of great outflow of lava supports a Middle Keweenawan age
for these dikes. Dikes which may have fed some of these
flows intersect the Lower Keweenawan sequence in the Go-
gebic area (Irving and Van Hise 1892). Comparison of the
Central Mine group flows with the Gogebic dikes (Table #3)
confirms their close chemical relationship with one another.
The dikes could have intruded at any dip. If they did in-
trude at the present dip they were probably emplaced at the
end of Keweenawan time. However, late Keweenawan intru-
sives in Michigan were more feldSpathic than the olivine
diabase dikes of the Gogebic range.

The olivine diabase dikes of northern Michigan appear
to fit the hypothetical stress pattern of a subsiding linear
basin rather than the hypothesis of compressional release
and intrusion. The hypothesis of a subsiding linear basin
accounts for all of the known facts without having to call
upon special conditions. This conclusion would place the
olivine diabse dikes in the Keweenawan series, thus elimi-
nating a post-Animikie - pre-Keweenawan age. An upper
limit in time would be placed at pre-Jacobsville. If cor-

relation of the dikes of the Gogebic area with those in the

43

LMarquette district is accepted, a lower time limit of Middle
IKeweenawan is placed on them, but as was mentioned in the

section on petrography, these dikes may not have been in-

truded at the same absolute time.

10

FIGURE

 

 

2520093

 

2 <9 _ I u. 2
I 111/
3.5 :3 .3 3.3m moEmanm <.._.ommzz_s.
ltflN..°—.~OBOO / w¥<l— ................ -..
# 'el. £x< Emom /
z <, 8
2...... o 25:20 m_x<

owl-MPH

es. . 2.93 mm... 0... ”1224mm
_ muxa $425 3.4.. .3 SEE

 

 

REVERSED POLARITY

In the northern hemisphere, magnetic formations ge-
nerally attract the north seeking pole of the dip needle.
However, most of the late Precambrian diabase dikes of
northern Michigan exhibit a reverse polarization which
attracts the south seeking pole. This polarization is
inclined approximately -85° to the present geomagnetic
field, and may have significance as a correlative tool.

Igneous rocks acquire most of their magnetism as they
cool through, and slightly below the Curie point of their
ferromagnetic minerals. Magnetic intensities for fields as
small as that of the earth (approximately 0.5 oersteds)
are normally prOportional to the field (Nagata 1953).
Thermo-remnent magnetism acquired in this manner is ge-
nerally assumed to lie accurately in the direction of the
ambient field. However, recent work by Strangway (1960)
has indicated that there is a tendency for some stable mag-
netism to align parallel to the strike of the dike. This
preferential alignment is apparently due to internal de-
magnetization that is dependent upon the geometric shape
of the body and its relationship to the geomagnetic field.
The quantity of magnetism so aligned is small but not neg-
ligible. This may necessitate the reevaluation of paleo-
magnetic data previously obtained from dikes.

Isothermal magnetization can generally be neglected
in paleomagnetic studies, as the earth's magnetic field

is small compared to the coercive forces of rock minerals

45

46

(more than 100 oersteds, Nagata 1953). However, lightning
‘with large magnetic fields may affect limited areas of the
rock. Magneto-crystalline anisotropy is generally neglected
in paleomagnetic study of undeformed hypabyssal igneous
rocks because of random orientation of the crystals prior
to the Curie temperature. Magnetostriction may be impor-
tant in its ability to alter the magnetic components of
minerals (Graham et al., 1957) but as yet, insufficient
data is available to evaluate it effectively. This pro-
perty may also be time dependent (Graham 1959) which would
further complicate its evaluation.

From the preceding it follows that igneous rocks
generally portray the ambient magnetic field at the time
of solidification, providing no physical (rotation, in-
version, folding), chemical (alteration, replacement) or
a combination of these processes has taken place. The re-
verse polarization observed in the olivine diabase dikes
in northern Michigan can then be stated to probably be due
to one of two causes: either the rock cooled in a geomag-
netic field similar to the present magnetic field and was
reversed due to subtle physical-chemical changes as dis-
cussed by Neel (1951, in Blackett 1956), Neel (1955), Gra-
ham (1953), Nagata (1953) and others, or the rock cooled in
the direction of the geomagnetic field which at that time
was essentially the reverse of the present sense (Runcorn
1954, 1955. and others).

Graham (1953) studied several reversely polarized late

47

Precambrian olivine diabase dikes in Baraga county, Michi-
gan. To account for the reversal, he revised a hypothesis
of Neel (1951, in Blackett 1956). Neel called for two com-
ponents of different Curie points: component A with a higher
Curie point imposing a field, which would be opposite on
component B with the lower Curie point during the cooling
process. Selective chemical action then removes component
A and leaves component B with the reversed magnetic moment.
Graham hypothesises that similar results could be obtained
if only one component was present. His proposal was that
one constituent A (magnetite), dispersed by essentially non-
magnetic ilmenite lamellae, could be oxidized to a second
constituent B (maghemite) at the expense of smaller par-
ticles of A. These small particles of constituent B would
be magnetized in the inverse sense by the external field

of the first magnetized constituent A in the manner sug-
gested by Neel. The external moment of the sample then
would later become reversed as a result of demagnetization
in time of the larger particles of constituent A (magnetite)
which would be too large to remain as domains for long
periods of time and could be expected to be demagnetized

by forming domain field closure loops with themselves. The
smaller size volume of constituent B (maghemite) with the
inverse moment, will remain as stable domains with a high
coercive force. The chief function of the ilmenite is to
break up the large magnetic grains so they can behave like

an aggregate of small particles whose fields interact

48

(Graham 1953). Graham also found that a significant por-
tion of the ferromagnetic material present in the dikes
'was not contributing to the external natural remanent
moment.

A polished section and a qualitative chemical analy-
sis by the writer confirmed the presence of ilmenite
lamellae in the magnetic fraction of the rock. This tends
to support Graham's hypothesis, however, the hypothesis
should be dependent upon the identification of the maghemite
in the rock.

A modification of Graham's (1953) hypothesis, which
does not require the formation of maghemite, is suggested
in the light of work by Strangway (1960). Strangway be-
lieves that some of the magnetic moment of the dikes is the
result of magnetite located in slightly altered olivine
and pyroxene crystals(Figure #11). This powder magnetite
is presumed to have been formed by deuteric alteration.
Powder magnetite might be affected in a manner suggested
by Neel (1951, in Blackett 1956) by assuming a polarization
opposite the field of the previously formed magnetite.

The first formed, larger magnetite, would demagnetizs in
time as suggested by Graham, thus giving a residual mag-
netic inversion. Strangway found the stable magnetism,
which he attributed to the powder magnetite, to lie ge-
nerally in the plane of the dike. However, this is depen-
dent upon the geometric shape of the body and its orienta-

tion in the magnetic field.

49

The minute size of this later formed powder magnetite
suggests great stability, and because of its size the Curie
point would be lowered (Neel 1955). Lowering of the Curie
point would remove a strong objection to Graham's and simi-
lar hypotheses, put forth by Nichols (1955) in which he
felt that because maghemite must form from magnetite at
temperatures below 600° C and consequently, below its Curie
point (675° C) it would be magnetically unstable, whereas
Graham and others have found the magnetism of the dikes to
be quite stable.

In addition to the above methods of reversal, several
other chemical models have been suggested (Neel 1951, in
Blackett 1956) some of which have been confirmed as con-
tributing to reversal of polarization. There is however,
no singularly clear method which explains all of the re-
versals. Additionally, an observation which would question
chemical reversal might be the chemical similarity between
the Keweenawan flows and the fresh dikes, since there are
no known reversals of polarization in the Keweenawan lavas.

The alternative hypothesis of dipole reversal eXplains
quite simply all reversed polarizations without calling
upon special physical-chemical conditions. However,sev-
eral observations suggest that the dikes were reversed by
some method other than dipole reversal. The varied direc-
tions of magnetic orientation found by Graham (1953), the
finding of a dike that was normally polarized on one side

and reversely polarized on the other by Strangway (1960),

50

Spiroff's (1962, personal communication) finding of dikes
that change in direction of polarity along strike, suggest

chemical-physical changes rather than geomagnetic inversion.

Figure 11

Sample #5, 40 power plane light. Olivine -
Pyroxene altering to magnetite. Possible
source of powder magnetite?

51

‘Figure 11

 

52

ECONOMIC IMPORTANCE OF THE DIKES

There have been many questions raised as to the age
and methods of enrichment of the soft ores of northern
Michigan. Royce (in Newhouse 1942) believes that the dike
in the Isabella mine, in conjunction with ground water,
was instrumental in concentrating the soft ore at this
location.

"This ore body is formed by three

horizons in the iron formation, which

are bent in a fold trough pitching flatly

to the northwest and cut across by the

Isabella dike. The combined trough in-

fluence of the fold and of the dike has

resulted in the concentration of ore in

three favorable horizons." "The concen-

tration of this ore body was by water cir-

culation down the trough and along the dike.”
If Royce is correct, then enrichment of the ore was
post dike intrusion. If the dikes of this study, of
which the Isabella is one, are Keweenawan in age, a
lower time limit can be placed on the enrichment, at least
in the Gogebic district and in the Isabella mine.

If most of the enrichment took place in post-Kewee-
nawan time, a rapid economical method of locating soft
ore bodies that are controlled by the late dikes could
be envisioned. As these dikes are reversely polarized
they are readily observed on both ground and aeromag-
netic surveys. Inspection of dike locations for trough
influence that could concentrate the ore would deliniate

areas for further exploration work. This method assumes

that enrichment was by ground water, which seems likely

53

when the position of the ore bodies relative to the

dikes is considered.

54

CONCLUSION

The fresh dikes of northern Michigan outcrop in at
least a 50 mile wide belt, which apparently extends from
the Gogebic district through Marquette into Lake Superior.
They are all similar to one another microscopically and
in field characteristics, and are believed to be of the
same system. These dikes intersect the folded Animikie
sediments in Marquette, Baraga, Gogebic and Iron counties;
and they probably intersect the post-Animikie - pre-Kewee-
nawan granite in Dickinson county. The late dikes in the
Gogebic district, thought to be equivalent of late dikes
in the other areas, intersect some of the Lower Keweenawan
lava flows. In Baraga county the dikes are overlain by
the Jacobsville sandstone of probable Lower Cambrian age.
The intrusion pattern indicates that the fresh dikes were
controlled mainly by Keweenawan stresses. The chemical
similarity between the fresh dikes and the Keweenawan lava
flows suggests a similar magma source. The writer there-
fore concludes that the late fresh dikes of northern

Michigan are Keweenawan in age.

55

REFERENCES CITED

Allen, R.C., 1914, Correlation and structure of the Pre-
.cambrian formations of the Guinn iron-bearing dis-
trict of Michigan: Jour. Geology, v. 22.

Anderson, E.M-, 1942, The dynamics of faulting and dike
formation, with application to Britain, Oliver and
Boyd: Edinburgh.

Balsley, J.R., James, H.L., and Wier, H.L., 1949, Aero-
magnetic survey of parts of Baraga, Houghton and
Iron counties, Michigan, with preliminary geologic
interpretation: U.S. Geol. Survey, GeOphys. Inv.
Prelim. Rept.

Bayley, W.S., 1904, The Menominee iron-bearing district
of Michigan: U.S. Geol. Survey Mon. 46, 513 p.
Bayley, R-W-, 1959, Geology of the Lake Mary quadrangle,

Michigan: U.S. Geol. Survey Bull. 1077.

Blackett, P.M.S., 1956, Lectures on rock magnetism:
Israel: Weiman Science Press.

Broderick, T.M-, 1935. Differentiation in lavas of the
Michigan Keweenawan: Geol. Soc. America Bull. 44,
Po 503-558.

Butler, 3.8., and Burbank, W-S-, 1929, The copper depo-
sits of Michigan: U.S. Geol. Survey Prof. Paper
144, 238 p.

Buzas, 4., 1960, The interpretation of the aeromagnetics
of southeastern Marquette county, Michigan: Michigan
State University, unpublished.

Clements, J.M., Smyth, H.L., Bayley, W.S., and Van Hise,
C.R., 1899, Crystal Falls iron-bearing district of
Michigan: U.S. Geol. Survey Mon. 36.

Cornwall, H.R., 1951, Differentiation in magmas of the
Keweenaw series: Jour. Geology, v. 59, p. 151-172.

Cooke, H.C., 1946, Problems of Sudbury geology, Ontario:
Geol. Survey Can. Bull. 3.

Gair, J.E., and Wier, K.L., 1956, Geology of the Kiernan
quadrangle Iron county, Michigan: U.S. Geol. Survey
Bull. 1044, 87 p.

Graham, J.W., 1953, Changes of ferromagnetic minerals
and their bearing on magnetic properties of rocks:
Jour. GeOphys. Research, v. 58, p. 243-260.

Graham, J.W., Buddington, A.F., and Balsley, J.R., 1957,
Stress-induced magnetizations of some rocks with
analyzed magnetic minerals: Jour. GeOphys. Research,
v. 62, P. 465-474.

Graham, J.W., 1959, Magnetostriction and paleomagnetism
of igneous rocks: Nature, v. 183, p. 1318.

Hefner, N., 1951, Stress distribution and faulting: Geol.
SOC. America 31111., V. 62, P. 373-398.

Hamblin, W.K-. 1958, Cambrian sandstones of northern
Michigan: Mich. Dept. of Conservation, Geol. Sur-
vey Div. Publ. 51, 146 p.

Hess, H.R., 1949, Chemical composition and optical proper-
ties of common clinopyroxene: American Mineralogist,
v. 34, p. 621-666.

56

Hotchkiss, w.0., 1923, The Lake Superior geosyncline:
Geol. Soc. America Bull., v. 34, p. 667-678.

Irving, R.D., and Van Hise, C-Ho, 1892, Penokee iron-
bearing series of Michigan and Wisconsin: U.S.

Geol. Survey, Mon. 19.

James, H.L., and Dutton, C.E., 1951, Geology of the
northern part of the Iron River district: U.S.

Geol. Survey Ciro. 120.

James, H.L., 1955. Zones of regional metamorphism in the
Precambrian of northern Michigan: Geol. Soc. America
Bull., v. 62 no. 3. p. 251-266.

1958, Stratigraphy of pre-Keweenawan rocks in
parts of northern Michigan: U.S. Geol. Survey Prof.
Paper 3140, 44 p0

James, H.L., Clark, L. D., Lamey, c.A., Pettijohn, F.J.,
1961, Geology of central Dickinson county, Michigan:
U.S. Geol. Survey Prof. Paper 310, 175 p.

Johannsen, A., 1939. A descriptive petrography of the
igneous rocks: University of Chicago Press, Illinois,

0 1.

Lane, A. C., 1911, The Keweenawan series of Michigan: Pub.
6 (Ge01.Ser. 4) 2 vol., 983p. Map.

Larsen, E S-, amd Barman, H., 1934, The microscopic de-
termination of the nonopaque minerals: U.S. Geol.
Survey Bull., 848, 266p.

Moore, 3.8., and Armstrong, H.S., 1946, Iron deposits
in the district of Algoma: Ontario Dept. of Mines,
v. 55, part 4, p. 1-118.

Moorhouse, W.W., 1959. The study of rocks in thin section:
New York: Harper and Brothers, 514p.

Nagata, T., 1953. Rock magnetism: Tokyo: Maruzen Co. Ltd.,
225p.

Neel, L., 1955. Some theoretical aspects of rock mag-
netism:,Advances in Physics, V. 4, p. 191- 243.

Newhouse, W-d , 1942 ed., Ore deposits as related to
structural features: Princeton University Press.

Nicholle, G-D-, 1955, Mineralogy of rock magnetism:
Advances in Physics, v. 4, p. 2- 190.

Niggli, P., 1954, Rocks and mineral deposits: San Fran-
cisco: W. H. Freeman 00., 559p.

Rominger, 0. Lo: 1881, Geology of Michigan; v. 5.

Runcorn, S.K., 1955, Rock magnetism - geOphysical as-
pects: Advances in Physics, v. 4, p. 244-291.

Sanford, A.R-. 1959, Analytical and experimental study
of simple geologic structure: Geol. Soc. America
B111]... V. 70, p0 19’520

Strangway, D.W., 1961, Magnetic properties of diabase dikes:
Jour. GeOphys. Research, v. 66, p. 3021-3032.

Thiel, E., 1956, Correlation of Gravity anomalies with the
Keweenawan geology of Wisconsin and Minnesota: Geol.
Soc. America Bull., v. 67.

Thomson, C.E., 1956, Geology of the Sudbury basin:
Ontario Dept. of Mines, v. 65, part 3, p. 1—56.

57

Van Hise, O.R., and Bayley, W.S., 1897. Marquette iron-
bearing district of Michigan: U.S. Geol. Survey,
Mon. 28. __

Van Hise, C-R', and Leith, 0.4., 1911, Geology of the Lake
Superior region: U.S. Geol. Survey, Mon. 52.

Zurbrigg, H.F., 1961, (personal communication).

 

 

 

 

   

.1114111411111111~1