Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Geology of the northern end of Juan de Fuca Ridge and adjacent continental slope Barr, Sandra Marie 1972

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1973_A1 B37.pdf [ 29.3MB ]
Metadata
JSON: 831-1.0052648.json
JSON-LD: 831-1.0052648-ld.json
RDF/XML (Pretty): 831-1.0052648-rdf.xml
RDF/JSON: 831-1.0052648-rdf.json
Turtle: 831-1.0052648-turtle.txt
N-Triples: 831-1.0052648-rdf-ntriples.txt
Original Record: 831-1.0052648-source.json
Full Text
831-1.0052648-fulltext.txt
Citation
831-1.0052648.ris

Full Text

GEOLOGY OF THE NORTHERN END OF JUAN DE FUCA RIDGE AND ADJACENT CONTINENTAL SLOPE by SANDRA MARIE BARR B.Sc. University of New Brunswick, 1968 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of Geology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December, 1972 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make i t freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Geological Sciences The University of British Columbia Vancouver 8, Canada Date December 11. 1972 ABSTRACT Bathymetry, seismic reflection profiles, magnetic data, and dredged basalt indicate that two centres of spreading have existed near the northern end of Juan de Fuca Ridge in the last 0.7 m.y. (Brunhes Geomagnetic Polarity Epoch). A short-lived eastern branch functioned only in the early part of the Brunhes Epoch. More prolonged and recent activity has occurred on the western branch, and has involved extrusion of basalt, rifting, and uplift. The anomalous relief of the northern end of Juan de Fuca Ridge compared to the ridge further south is the result of a tectonic event which occurred 100,000-350,000 years ago. Prior to this, turbidites were deposited over most of the northern part of the ridge, which consisted of a series of basement ridges and valleys of subdued relief trending NNE parallel to the magnetic anomalies. During the event, the igneous basement together with the overlying turbidites was faulted, tilted, and uplifted along the primary structural trends. The basement relief was increased in places to more than 2 km. Subsequent turbidite deposition has been confined to valleys. The Sovanco Fracture Zone between Juan de Fuca and Explorer Ridges also suffered uplift. The tectonic disturbance may have resulted from an increase in spreading rate or the change in location of the axis of spreading on Juan de Fuca Ridge, or compressional interaction between Explorer sub-plate to the north and Sovanco Fracture Zone and the northern end of Juan de Fuca Ridge. Sovanco Fracture Zone, consisting of a WNW-trending basement ridge (Sovanco Ridge) with complex associated mag-netic anomalies, may have originated as a strike-slip fault or fault-zone trending east-west about 5 m.y. ago. Location of present transform movement within the fracture zone is equivocal, but apparently trends NW. A specimen of basalt from the eastern end of Sovanco Ridge yielded a fission track age of <309,000 years. Basalts dredged from 13 locations on Juan de Fuca Ridge, Sovanco Ridge, and Heck and Heckle Seamount Chains on the western flank of Juan de Fuca Ridge are low-potassium tholeiites of the ocean ridge type. Twenty-seven representative specimens of fresh basalt, as well as 15 samples of glass selvages and altered basalt, were analyzed for 9 major elements by an atomic absorption technique set up by the author. The analyses are closely similar, showing no well-developed differentiation trend. They straddle the boundary between the olivine-normative and quartz-normative fields, and the basalts may have been derived by segregation from the mantle at a depth of about 15 km. A specimen from NE of the eastern spreading centre yielded a fission track age of 0.95 ± 0.53 m.y. which is in reasonable agreement with the age of 0.7-0.9 m.y. predicted by the magnetic anomaly time scale. Increasing thicknesses of Fe-Mn coatings on dredged basalts indicate that the two NW-trending seamount chains increase in age away from the crest of Juan de Fuca Ridge. The seamounts originated and grew to virtually their present sizes while s t i l l at or near the ridge crest. Reflection profiles across the continental slope indicate that i t has formed by folding, reverse faulting, and uplift of strata of Cascadia Basin. Internal deformation has caused these sediments to behave as acoustical basement in the structures beneath the slope but their sedimentary nature has been confirmed by dredging. The presence of magnetic anomalies 5-8 m.y. old beneath the lower and middle continental slope indicates subduction. The magnetic basement dips below the slope at an angle of more than 1 0°. iv TABLE OF CONTENTS ABSTRACT i i LIST OF TABLES v i i i LIST OF ILLUSTRATIONS x ACKNOWLEDGMENT *v I. INTRODUCTION LOCATION 1 PURPOSE OF INVESTIGATION 1 PREVIOUS WORK 3 II. THE GEOLOGICAL FRAMEWORK REGIONAL GEOLOGY AND GEOPHYSICS 6 Physiography 6 Magnetics 9 Heat Flow 12 Seismicity 14 Gravity 16 Seismic Refraction 19 Sea Floor Geology 19 REGIONAL PLATE TECTONICS 22 OCEANIC BASALTS 27 Composition of Oceanic Basalts 29 Origin of Oceanic Basalts 30 V III. STRUCTURAL FEATURES OF THE OCEAN FLOOR IN THE STUDY AREA BATHYMETRY 36 Bathymetric Chart Preparation 36 Physiography 40 Continental Slope 40 Cascadia Basin 41 Juan de Fuca Ridge 41 Sovanco Fracture Zone 43 Seamount Chains 43 Abyssal Plains 44 SEISMIC REFLECTION DATA 45 Introduction 45 Geologic Units 47 Juan de Fuca Ridge Area 47 Igneous Basement 47 Sedimentary Sequence 50 Continental Slope 58 Sedimentary "Basement" 59 Stratified Continental Slope Unit 59 Pleistocene Wedge 59 Structural Interpretations 60 Juan de Fuca Ridge 60 East Ridge 61 East Valley 62 Middle Ridge 63 Middle Valley 64 West Ridge 65 West Valley 66 Ridge Crest Near 48°N Latitude 67 Summary 67 Sovanco Fracture Zone 68 Seamount Chains 70 Abyssal Plains 71 Cascadia Basin 76 Continental Slope 78 Summary of Structural Interpretations 83 vi MAGNETIC DATA 84 Introduction 84 Juan de Fuca Ridge 85 Sovanco Fracture Zone and Cascadia Basin to the North 89 Abyssal Plains 89 Seamount Chains 89 Continental Slope and Eastern Cascadia Basin 90 Discussion and Summary of Magnetic Data 94 IV. SAMPLES FROM THE OCEAN FLOOR INTRODUCTION 96 DREDGING TECHNIQUE 96 LABORATORY INVESTIGATIONS 100 BASALTS AND RELATED DEPOSITS 101 Petrography of the Basalts 101 Juan de Fuca Ridge 104 Sovanco Ridge 108 Heck and Heckle Seamount Chains 112 Significance of the Petrography 115 Porphyritic basalts 115 Faulting 118 Mode of Eruption 118 Chemistry of the Basalts 119 Juan de Fuca Ridge 123 Sovanco Ridge 128 Seamount Chains 130 Comparative Discussion of the Chemical Results 131 Petrogenesis of the Basalts 134 Chemical Analyses of Basaltic Glass 146 Alteration of the Basalts 149 Altered Rims 150 Irregular Alteration 154 Hyaloclastites and Aquagene Tuffs 155 Hyaloclastites 155 Aquagene Tuffs 156 Ferro-manganese Deposits 160 Ages of the Basalts 162 Ferro-manganese Geochronology 163 Fission Track Age Dating 165 Implications of the Relative Ages of the Basalts 169 vii SEDIMENTARY DEPOSITS 170 Introduction 170 Juan de Fuca Ridge and the Seamount Chains 171 Semi-consolidated Deposits 171 Unconsolidated Deposits 171 Continental Shelf 172 Unconsolidated Deposits 172 Consolidated Deposits 172 Conclusions 173 V. DISCUSSION: RECENT GEOLOGICAL HISTORY OF THE 175 NORTHERN END OF JUAN DE FUCA RIDGE AND RELATED FEATURES Spreading Centre 176 Physiography 176 Seismic Reflection Data 179 Magnetic Data 182 Characteristics of Dredged Basalts 183 Seamount Chains 184 Sovanco Fracture Zone 186 Uplift and deformation at the northern end of Juan de Fuca Ridge 191 The Subduction Zone 196 VI. SUMMARY AND CONCLUSIONS 203 BIBLIOGRAPHY 211 APPENDIX A: Seismic Reflection Profiles 222 APPENDIX B: Part 1 Petrographic Descriptions of Basaltic Rocks 245 Part 2 Whole-Rock Chemical Analysis of Basalt by Atomic Absorption after Fusion with Lithium Metaborate 268 v i i i LIST OF TABLES I. Depths to sea floor, igneous basement, and the top of Unit A at abyssal plain and associated locations. 74 II. Major petrographic characteristics of principal types of basalt dredged from Stations 1 to 13 102 III. Partial chemical analyses and normative compositions of fresh basalts from Juan de Fuca Ridge and Sovanco Ridge 120 IV. Partial chemical analyses and normative compositions of fresh basalts from Heck and Heckle Seamount Chains 121 V. Recalculated partial chemical analyses of porphyritic basalts from Juan de Fuca Ridge and Sovanco Ridge. 125 VI. Mean partial chemical analyses and standard deviations of non-porphyritic and porphyritic basalts from Juan de Fuca and Sovanco Ridges, and of basalts from the seamount chains 132 VII. Partial chemical analyses of fresh basalts and their glass selvages--Juan de Fuca Ridge, Heck Seamount 147 VIII. Summary of major types of alteration observed in basalts from the study area 151 IX. Partial chemical analyses of fresh interiors and altered rims of basalts from Juan de Fuca Ridge, Heck Seamount Chain and Heckle Seamount Chain 152 X. Minimum U contents of basalts from the study area, determined from fission track concentrations in irradiated specimens 166 XI. Fission track data for two specimens from Juan de Fuca and Sovanco Ridges 167 XII. Seismic reflection profiles 222 XIII. Summary description of dredge hauls in which samples of underlying bedrock were obtained 262 XIV. Description of dredge stations at which no samples of underlying bedrock were obtained 265 ix XV. Analytical conditions for atomic absorption measurements 274 XVI. Comparisons between partial chemical analyses of basalts analyzed by various laboratories and by the author. Mean deviations of four aspirations are presented for each oxide of the author's analyses 278 XVII. Maximum mean deviations among the four aspirations of each sample recorded within each sample batch. 280 XVIII. Comparisons between replicate partial chemical anal-yses in the four batches of samples. Mean deviations of four aspirations are also presented for each oxide of the author's analyses 281 XIX. Petrographic descriptions of analyzed specimens ' 282 X LIST OF ILLUSTRATIONS FIGURE 1. Bathymetric chart of the sea floor west of Vancouver Island, Washington, and Oregon. The study area is outlined 2 2. Physiographic provinces for oceanic regions from off northern California to British Columbia 8 3. Raff and Mason magnetic anomaly pattern in the north-east Pacific Ocean 10 4. Magnetic anomalies of Raff and Mason (1961) superimposed on bathymetry 13 5. Map of epicentres of earthquakes occurring between 1954 and 1963 15 6. Epicentres of earthquakes (1954-1969) superimposed on bathymetry 17 7. Free-air gravity anomalies superimposed on bathymetry 18 8. Boundaries of the Juan de Fuca Plate 23 9. Reconstruction of plate evolution and deformation related to Cenozoic interactions of the North American and Pacific plates and the North American and Kula plates 25 10. Locations of plate boundary regimes with respect to points in western North America 26 11. A to E. Schematic model of plate interactions. F. Evolution with time of boundary regimes along the coast of North America 28 12. Model of magmatic evolution for the Mid-Atlantic Ridge 32 13. Ship tracks from which were obtained bathymetric data used in preparation of revised bathymetric chart of northern end of Juan de Fuca Ridge 38 14. Ship tracks from which were obtained bathymetric data used in preparation of revised bathymetric chart, continental slope area 39 xi FIGURE 15. Revised bathyraetric chart of the northern end of Juan de Fuca Ridge and adjacent continental slope 287 16. Geological map of the Juan de Fuca Ridge area north of 48°N latitude 48 17. Depth below sea level to igneous basement in seconds of 1-way travel time 49 18. Minimum sediment thickness over the Juan de Fuca Ridge area north of 48°N latitude 51 19. Locations of seismic profiles 54 20. Line-drawing interpretations of seismic reflection profiles to illustrate the distributions and interrela-tions of sedimentary units A and B 55 21. Line-drawing interpretations of seismic reflection profiles to illustrate the distributions and interrela-tions of sedimentary units A and B 56 22. Fence diagram (isometric projection) of interpretations of seismic reflection profiles to show structure and distribution of geologic units near the northern end of Juan de Fuca Ridge 57 23. a. Line-drawing interpretation of seismic reflection profile L-L'. b. Line-drawing interpretation of seismic reflection profile M-M', with coincident magnetic anomaly profile 69 24. Locations of points used in Table I to illustrate depths to sea floor, igneous basement, and the top of Unit A on Juan de Fuca Abyssal Plain and adjacent features 73 25. Comparisons between residual magnetic anomaly profiles obtained with seismic reflection profiles D-D' and E-E', and the Raff and Mason magnetic anomalies along the same lines 86 26. Magnetic anomalies under eastern Cascadia Basin and the continental slope between 48° and 49°N latitude contoured from magnetic data recorded during the present investigation 91 27. Magnetic anomaly source depths under eastern Cascadia Basin and the continental slope between 48° and 49°N latitude, calculated by the graphical "straight-slope" method 92 x i i FIGURE 28. Locations of dredge stations in the Juan de Fuca Ridge area in which samples of bedrock (basalt) were recovered 97 29. Locations of dredge stations on the continental slope in which samples of bedrock (mudstone) were obtained 98 30. Basalts from Juan de Fuca Ridge (Middle Ridge) 105 31. Basalts from Juan de Fuca Ridge (West Ridge and Endeavour Seamount) 106 32. Basalts from the eastern end of Sovanco Ridge 109 33. Photomicrographs of basalt from the eastern end of Sovanco Ridge, illustrating the typical transition from glass selvage to crystalline interior observed in basalts from the study area 110 34. Basalts from the southern flank of a peak on Sovanco Ridge 111 35. Basalts from Heck Seamount Chain 113 36. Basalts from Heck and Heckle Seamount Chains 114 37. Alkali/silica diagram of the basalts from the study area 135 38. Oxide concentrations vs. S. I. for seamount chain basalts 137 39. Oxide concentrations vs. S. I. for ridge basalts 138 40. (Na20 + K20)-Fe0 (total iron)-Mg0 diagrams for basalts from the study area and other basalts and gabbros 139 41. a. FeO against MgO and b.. CaO against A I 2 O 3 for basalts from the study area 141 42. a. Normative mineralogy of basalts from the study area, plotted on a nepheline-olivine-anorthite-hypersthene-quartz diagram, b. Fields of various submarine basalts on normative mineralogy diagram as in a. 143 43. Hyaloclastites and tuffs from Juan de Fuca Ridge and Heck and Heckle Seamount Chains 157 44. Tuffs and ferro-manganese deposits from Heck and Heckle Seamount Chains 159 x i i i FIGURE 45. Maximum thicknesses of ferro-manganese coatings on rocks from each dredge station plotted against the position of each station projected onto a line through Station 4 perpendicular to the magnetic anomalies 164 46. Magnetic anomalies of Raff and Mason (1961) superimposed on bathymetry near the northern end of Juan de Fuca Ridge 177 47. Postulated axis of most recent sea floor spreading near the northern end of Juan de Fuca Ridge 178 48. a. Topography and bottom character of Escanaba Trough and walls on Gorda Ridge, b. Schematic interpretation of the structure of Juan de Fuca Ridge near its northern end 181 49. a-c. Postulated mode of adjustment of a spreading centre to a change in spreading direction, d-f. Postulated sequence of events in the development of the northern part of Juan de Fuca Ridge in the last 8 million years 187 50. a. Vectors diagrams for plate motions in the northeast Pacific, b. Schematic block diagram of continental slope showing interpretation of structures caused by compressive motion between Juan de Fuca and American Plates 197 51. Locations of seismic reflection profiles in the area of the northern end of Juan de Fuca Ridge reproduced in Plates I to XII 223 52. Locations of seismic reflection profiles over the continental slope reproduced in Plates XIII to XIX 224 53. Key to seismic reflection profiles, Plates I-XIX and Figure 23 225 xiv PLATE SEISMIC REFLECTION AND MAGNETIC PROFILES I. A-A» Juan de Fuca Ridge 226 II. B-B' Juan de Fuca Ridge 227 III. C-C* Juan de Fuca Ridge 228 IV. D-D' Juan de Fuca Ridge 229 V. E-E* Juan de Fuca Ridge 230 VI. F-F1 G-G' Juan Juan de de Fuca Ridge Fuca Ridge (Middle Ridge) (Middle Ridge) 231 VII. G-H H-H* Juan de Sovanco Fuca Ridge (Middle Valley) Fracture Zone 232 VIII. I-11 Sovanco Fracture Zone to Heck Seamount 233 IX. J-J" Sovanco Fracture Zone 234 X. J-K' Sovanco Fracture Zone 235 XI. K-K' Sovanco Fracture Zone 236 XII. N-N' South from the eastern end of Heck Seamount Chain 237 XIII. O-O' p_p< Lower continental Lower continental slope slope 238 XIV. Q-Q' Continental slope 239 XV. R-R1 Continental slope 240 XVI. S-S1 Continental slope 241 XVII. T-T* Lower continental slope 242 w i n . U-U' Lower continental slope 243 XIX. V-V Continental slope 244 XV ACKNOWLEDGMENTS. This thesis was done under the supervision of Dr. R. L. Chase to whom the author is very indebted. Thanks are due to Dr. W. K. Fletcher for advice concerning chemical analyses of basalts, and to Dr. E. G r i l l , Institute of Oceanography, U. B. C , for permitting use of his laboratory and atomic absorption f a c i l i t i e s . The advice and assistance of the technicians and staff of the Department of Geology, U. B. C , are appreciated. Special thanks are due to R. MacDonald for help in organizing cruises. I wish to thank the captains, officers, and crews of the C.F.A.V. ENDEAVOUR and C.S.S. PARIZEAU for their invaluable co-operation and assistance. The enjoyable weeks spent aboard those vessels will always be remembered with pleasure. Dr. F. Aumento, Dalhousie University, instructed the author in the techniques of fission track age dating and arranged for specimen irradiation. I wish to thank Dr. D. A. McManus, University of Washington, for collabora-tion and discussion, and for making available seismic reflection data from the study area. A. G. Thomlinson provided assistance in thought and moral support. Dr. B. Cameron kindly did microfaunal studies for the author. Financial support was provided by a National Research Council 1967 Science Scholarship held by the author from 1968-1972. Additional financial support was provided by NRC Research Grants to R. L. Chase and J. W. Murray, DRB Grant 9511-95 to R. L. Chase, a Department of Energy, Mines and Resources contract to J. W. Murray, and grants from the following o i l companies: Shell Canada Limited; Chevron Standard Limited; Imperial Oil Limited; Mobil Canada Limited; and Union Oil Company of Canada Limited. FRONTISPIECE CHAPTER I INTRODUCTION LOCATION Juan de Fuca Ridge is postulated to be an active segment of the world-wide system of spreading ocean ridges (Wilson, 1965; Vine and Wilson, 1965). It forms part of the boundary between the Pacific Lithospheric Plate and the small Juan de Fuca Lithospheric Plate west of southern Vancouver Island, Washington, and Oregon. The study area, between 48°N and 49°N latitudes, from the upper continental slope westward to 130°W longitude, is an area of about 32,000 square kilometres which includes the crest and flanks of the northern tip of Juan de Fuca Ridge, the Sovanco Fracture Zone, and the adjacent continental slope (Fig. 1). PURPOSE OF THE INVESTIGATION Juan de Fuca Ridge has been interpreted to be a continuation of the East Pacific Rise (Menard, 1960; Wilson, 1965), and Juan de Fuca Plate a remnant of the larger Farallon Plate (Atwater, 1970) which has been over-ridden by the North American Plate. Considerable speculation and discussion have concerned the recent history and tectonics of Juan de Fuca Ridge (Dehlinger et a l , 1970) and what is occurring at the Juan de Fuca Plate margins (Silver, 1969, 1971a, 1971b). However, when this project was begun (1969), few detailed geological and geophysical data were available in the area west of Vancouver Island. 2 FIGURE 1. Bathymetric chart of the sea floor west of Vancouver Island, Washington, and Oregon (after Mammerickx and Taylor, 1971). The study area is outlined. Contour interval 100 or 200 m, 25 in where significant. 3 Hence this study was undertaken to determine the structure and recent geological history of the area around the northern end of Juan de Fuca Ridge and to investigate tectonic activity there and under the continental slope to the east. In order to tackle these problems it was necessary to do a reconnaissance geological mapping of the area. This was accomplished during seven cruises when bathymetric data, continuous seismic reflection profiles, magnetic profiles, and sea floor samples were obtained. This thesis is primarily concerned with the analyses of these data, and the applications of the results to the problems under consideration. The bathymetric data have been combined with other bathymetric data available from the area, and a revised bathymetric chart produced. The seismic profiles have been interpreted and, in conjunction with the bathymetry and magnetic data, enable an interpretation of the structural elements and the tectonics. Rock samples were dredged from as many positive bathymetric features as possible (limited by time available and unwillingness of the sea floor in many places to give up samples) to determine the rock types under-lying the area. Basaltic rocks obtained are compared petrographically and chemically to one another and to similar rocks from other spreading ridges, fracture zones, and seamounts. These data enable postulation of a model for magmatic evolution beneath the study area. Finally, this information is synthesized in a discussion of the recent geological history and plate tectonics in the study area. PREVIOUS WORK . A limited amount of work has been done in the area as part of larger regional investigations which are described in Chapter II. The 4 purpose of this section is to emphasize that work which directly pertains to the study area. The first detailed bathymetric chart of the Juan de Fuca-Gorda Ridge region, which included the study area, was published by McManus (1964). In discussing the physiography of the region, McManus (1967) noted that crossings of the crest of Juan de Fuca Ridge in the present study area record a region 15-30 km wide of 175- to 250-fathom hills rising from a sea floor level of close to 1400 fathoms. He also described a flat-floored trough or median valley on the crest of Juan de Fuca Ridge in the same area. He con-cluded that the northern end of Juan de Fuca Ridge has more relief than the rest of the ridge to the south. A continuous seismic reflection profile near the northern end of Juan de Fuca Ridge (Hamilton, 1967) demonstrated that the volcanic basement rises westward from below the flat sediment surface of Cascadian Basin to crop out as ridges, before descending beneath the plain west of Juan de Fuca Ridge. Ewing et al (1968) published a continuous seismic profile across Juan de Fuca Ridge slightly to the north of Hamilton's profile. They noted that their profile appears to confirm the existence of Juan de Fuca Ridge crest at the position indicated by the magnetic anomaly pattern (Raff and Mason, 1961) and clearly shows a f i l l e d marginal trench at the foot of the continental slope. The profiles of Ewing ejt al_ and Hamilton both show con-siderable basement relief on the crest of Juan de Fuca Ridge, particularly a deep sediment-filled graben, which is the flat-floored median valley noted by McManus (1967). On the basis of the magnetic anomalies in the region (Raff and Mason, 1961), Pavoni (1966) identified a left-handed dislocation zone southwest of Vancouver Island, which he called the Sovanco Fault Zone (after South Of 5 VANCOuver Island). Other workers (Wilson, 1965; Vine and Wilson, 1965) have considered this to be a dextral transform fault. In either case, i t forms the northern termination of Juan de Fuca Ridge, and is herein referred to as the Sovanco Fracture Zone. The detailed bathymetric chart of the region (Mammerickx and Taylor, 1971) shows considerably more detail in the study area than McManus' 1964 publication, particularly the linear seamount chains on the western flank of Juan de Fuca Ridge and the complex features of the Sovanco Fracture Zone. McManus ejt al_ (1972) published an interpretation of a seismic reflection profile survey which included the north-central part of the study area. They concluded that the basement relief of the crest of Juan de Fuca Ridge and Sovanco Fracture Zone was formed after the time of creation of the magnetic anomalies associated with sea floor spreading, and that, although abruptly initiated, these vertical movements have continued in moderated form to the present. CHAPTER II THE GEOLOGICAL FRAMEWORK REGIONAL GEOLOGY AND GEOPHYSICS Much geological and geophysical survey work has been done in the region west of British Columbia and the northwestern United States, most of i t since 1960. To determine the geology and geological history of the study area, the relevant survey data and their interpretations for the region as a whole need to be considered. The purpose of this section is to present a brief summary of the data that have been collected and the interpretations that have been published to provide a framework for interpretations of data from the study area. Physiography Regional bathymetry was fir s t described by Menard and Deitz (1952) as part of the Ridge and Trough Province. Hurley (1960) and Menard (1964) further delineated and named the major physiographic provinces of the North-eastern Pacific Ocean. McManus (1964, 1965, 1967) published and discussed a detailed bathymetric chart of the area from off northern California to off northern Vancouver Island. Mammerickx and Taylor (1971) published a detailed contour map of the region from off southern Oregon to the Queen Charlotte Islands, part of the area surveyed by the U. S. Coast and Geodetic Survey ship PIONEER. The detailed bathymetry of this region is also included in Chart number 4 of the series "Bathymetry of the North Pacific" by Chase, Menard, and Mammerickx (1970). 7 Dominating the region is a chain of features of high relief extending from the continental margin off northern California to the con-tinental margin northwest of Vancouver Island (Fig. 2). This is a portion of the world-wide ocean ridge system with its associated fracture zones. Chains of seamounts are associated with the western flanks of the ridges, particularly Juan de Fuca. East of the ridges are Cascadia and Gorda Basins, flanked on the east by the continental slope. Physiographically, Gorda, Juan de Fuca, and Explorer Ridges are very different. Gorda Ridge has a prominent median valley (Escanaba Trough) with two flanking ridges which give way laterally to small ridges and elongate abyssal hills and troughs. Juan de Fuca Ridge, with generally low topography and no median trough, consists of elongate, parallel hills and valleys. Hamilton (1967) referred to Juan de Fuca as more of a low rise than a well-developed ridge. In contrast to the subdued topography on the ridge it s e l f , several of the seamounts in the northwest-trending chains on the western flank of the ridge rise nearly to sea level; for example, Cobb Seamount, summit -34 m. Explorer Ridge has an axial valley (Explorer Trough) which, however, extends as a topographic feature for 70 km south of the intersection of Sovanco Fracture Zone with the ridge. The western flank of Explorer Ridge has several seamounts, some of which form linear chains trending northwest. Prominent is the large, irregularly shaped Explorer Seamount, situated on the western flank of Explorer Trough, approximately opposite the intersection by Sovanco Fracture Zone. East of Explorer, Juan de Fuca, and Gorda Ridges, Cascadia and Gorda Basins have a depositional topography of deep-sea fans and plains. Along much of the eastern edges of Gorda and Cascadia Basins are marginal ridges which mark the base of the continental slope. FIGURE 2. Physiographic provinces for oceanic regions from off northern California to British Columbia (after Dehlinger et a l , 1970). 9 West of northern California and Oregon the continental slope trends approximately north-south, and consists of steps or benches (Dehlinger et a l , 1970). West of Washington and southern Vancouver Island where the trend of the continental slope is changing from north-south to northwest-southeast, a continental borderland consisting of ridges and intervening sediment-filled troughs is developed. In this region the continental slope is crossed by numerous submarine canyons. Towards the northern end of Vancouver Island, the continental slope narrows and steepens, and at least the upper part is a fault scarp (Tiffin et a l , 1972). Northwest of Vancouver Island, the continental slope is also narrow and steep, and is the locus of the Queen Charlotte Fault Zone. Magnetics Raff and Mason (1961), Mason and Raff (1961), Emilia et al (1968), and Couch (1969) reported magnetic anomalies in the region. The Raff and Mason (1961) anomaly map, as interpreted by Vine (1968) and modified by Silver (1971c) is reproduced in Figure 3. The predominantly north- and northeast-trending anomalies were used by Vine and Wilson (1965) and Vine (1966) to postulate that Juan de Fuca and Gorda are spreading ocean ridges. A spreading rate of 5.8 cm/year was calcu-lated for Juan de Fuca Ridge on the basis of the magnetic reversal time scale. Talwani et al_ (1965) concluded that Juan de Fuca is a mid-ocean ridge whose eastern flank now lies under the continent. The magnetic anomaly pattern suggests that Juan de Fuca Ridge rotated from a north-south orientation to its present orientation between 7 and 4 million years ago, although the details of the reorientation are equivocal. Emilia et_ a_l (1968) extended the magnetic survey data onto the continental slope and shelf off Washington and Oregon. They reported that FIGURE 3. Raff and Mason magnetic anomaly pattern in the north-east Pacific Ocean, modified from Vine (1968)(from Silver, 1971). 11 two magnetic lineations from Raff and Mason's survey area extend, apparently undisturbed, from Cascadia Basin across the continental slope to the shelf off northern Washington. Wilson (1965) suggested that the Queen Charlotte Fault and the Blanco Fracture Zone are dextral transform faults. This interpretation was later supported by earthquake first motion studies (Tobin and Sykes, 1968). Vine (1966), on the basis of his interpretation of the magnetic anomalies, identified the short Explorer Ridge northwest of Juan de Fuca, thus inferring that Sovanco Fracture Zone is also a dextral transform fault. The anomaly pattern is complicated by numerous other faults which generally trend either northwest or northeast, and divide the area into small plates. Pavoni (1966), Peter and Lattimore (1969) and Peter and De Wald (1971) have suggested that these dislocations, including Blanco and Sovanco Fracture Zones, originated as strike-slip faults. Pavoni (1966) interpreted that the oceanic crust in the region belonged originally to one crustal unit which was later dissected by wrench faulting along Blanco and Sovanco Fracture Zones and the other faults suggested by the magnetic anomaly pattern. Peter and Lattimore (1969) and Peter and De Wald (1971), with a different interpretation of the fault pattern, reconstructed Gorda, Juan de Fuca, and Explorer Ridges to form a single, continuous north-south trending feature approximately 10 million years ago. They suggested that deformation involved rotation, synchronous strike-slip faulting, and overthrusting of small crustal plates caused by northeast-southwest compression, rather than a gradual realignment of the ridges under the influence of changes in direction of sea-floor spreading, as suggested by Menard and Atwater (1968, 1969). Once in i t i a l displacement had occurred, active growth of individual 12 ridges would produce relative motion of the transform type on Blanco Fracture Zone and Sovanco Fracture Zone (called Explorer Fault by Peter and De Wald, 1971). Silver (1971c) suggested that synchronous activity of all faults on the magnetic map is not necessary. He showed that plate motions in the northeastern Pacific were complicated at about 2.5 m. y. B.P. by movement along the major northeast-trending fault Cj (Fig. 3), which is presently inactive, cutting Cascadia Basin. This movement would probably have resulted in a greater rate of net subduction along the continental margin off Oregon than off Washington and Vancouver Island. The magnetic anomaly pattern in the vicinity of the northern end of Juan de Fuca Ridge, Sovanco Fracture Zone, and Explorer Ridge departs from the trend and linearity of the rest of the region. A map of magnetic anomalies superimposed on bathymetry in this area (Fig. 4) shows prominent westerly-or northwesterly-trending anomalies between Juan de Fuca and Explorer Ridges, and irregular anomalies near the northern end of Juan de Fuca Ridge, near the southern end of Explorer Ridge, and in the region north of Juan de Fuca Ridge. It seems most likely that the rotation and internal deformation within Juan de Fuca Lithospheric Plate as indicated by the magnetic anomaly pattern are somehow related to the tiny size of the plate relative to the adjacent American and Pacific Plates. Heat Flow Heat flow data from the region were summarized by Dehlinger et_ al (1970), and additional data presented by Korgen ejt al (1971) and Lister (1970 a, b). Generally higher than normal heat flow values characterize the ridges. The heat flow profile across Juan de Fuca Ridge is asymmetrical. Unusually FIGURE 4. Magnetic anomal ies of Raff and Mason (1961) s u p e r i m p o s e d on ba thymet ry (Mammer ickx and Taylor, 1971). Time scale as interpreted by Vine (1968). 14 low values predominate on the western flank, whereas the eastern flank has high heat flow smoothly decreasing with distance from the crest. The latter profile is similar to that predicted by theoretical models, probably because the sediment cover on the eastern flank prevents convective water flow by sealing off porous crustal rocks from sea water. The western flank has much thinner sediment cover, and the heat flow is low, probably because of convective water flow in the crustal rocks (Lister, 1970b; Hyndman and Rankin, 1972). Seismicity In a compilation of earthquake data for the Northeastern Pacific, Tobin and Sykes (1968) showed a concentration of earthquake epicentres along Mendocino Fracture Zone, Gorda Ridge, Blanco Fracture Zone, and the Queen Charlotte Fault (Fig. 5), and with several focal mechanism solutions further substantiated the theory that Juan de Fuca and Gorda are spreading ridges by showing that directions of motion on the fracture zones are generally com-patible with transform faulting. The lack of seismic activity on Juan de Fuca Ridge compared to Gorda suggests that Juan de Fuca is more typical of the East Pacific Rise, much of whose crest is also aseismic. Earthquakes which do occur along the East Pacific Rise tend to concentrate along fracture zones between segments or ridge crest (Menard, 1967). Epicentres within Juan de Fuca Plate east of the Gorda and Explorer spreading segments suggest that these areas are presently deforming internally, local departures from rigid plate tectonics. In both these areas, the linear magnetic anomalies bend, and, in the Explorer area, are very irregular. Silver (1971b) found that faults in Gorda Basin which disturb young turbidites 15 FIGURE 5. Map of epicentres of earthquakes occurring between 1954 and 1963, reproduced from Tobin and Sykes, 1968 (from Dehlinger et a l , 1970). Earthquakes whose epicentres are shown were recorded by 10 or more seismograph stations. Stars indicate andesitic volcanoes of High Cascades-16 parallel the trends of the magnetic anomalies, suggesting later deformation of oceanic crust along lines of primary weakness. Although most of Juan de Fuca Ridge lacks seismic activity this is not strictly true of the ridge in the study area. A more detailed plot (Fig. 6) shows that a few earthquakes did occur on the eastern flank of the ridge between 1954 and 1969. However, most of the earthquakes were in the area north of Sovanco Fracture Zone and east of Explorer Ridge. The scattered line of epicentres south of the study area at about 47.5° is in the vicinity of an apparent offset in the crest of Juan de Fuca Ridge (Fig. 3). Gravity Free-air anomaly maps have been constructed for the region, and the data summarized by Dehlinger e_t al_ (1970). These show essentially zero-average free-air anomalies over the region as a whole and over major struc-tural features such as Juan de Fuca Ridge, inferring that they are in isostatic equilibrium. Many local anomalies are related to bottom topography; for example, the positive anomalies associated with seamounts and the negative anomalies associated with Explorer Trench (Fig. 7). Although the Scott Islands Fracture Zone is not a topographic trench, i t is a sediment-filled structure and hence has a large negative free-air anomaly. Similarly, in the study area, a negative anomaly is associated with the "median valley" on Juan de Fuca Ridge, which is also a deep sediment-filled structure. Negative anomalies, characteristic of oceanic-continental transition zones, are along the continental slope. Cascadia and Gorda Basins have only small anomalies, suggesting that there is no highly irregular basement topography in those Basins. 1961-1969 (from Seismioity of the United Statess USCGS, 1970) 01954-1961 (from Tobin and Sykes, 1968. EQ _ . . . recorded by 10 or more seismograph stations EQ recorded by 5 or more seismograph stations. FIGURE 6. Epicentres of earthquakes (1954-1969) superimposed on bathymetry (from Mammerickx and Taylor, 1971). Epicentres plotted to 50°N only. FIGURE 7. Free-air gravity anomalies (after Couch, 1969) superimposed on bathymetry (from Mammerickx and Taylor, 1971). 19 Seismic Refraction Refraction surveys by Shor et al (1968), Dehlinger et al_ (1968), and Couch (1969) show that Juan de Fuca Ridge is underlain by an up-raised crust and mantle, with the crust thinning over the ridge where i t overlies mantle of below-normal seismic velocity. Crustal structure of the ridge is apparently symmetrical. Gorda Ridge is similar. Comparative crustal sections are given by Dehlinger et al_ (1970). Sea Floor Geology Only a small number of studies of dredged rock samples have been made in the region. Engel and Engel (1963) reported that basalts from Mendocino Ridge and Gorda Escarpment are equigranular, have diabasic to gabbroic textures, and are predominantly tholeiitic in contrast to basalts dredged from tops and flanks of seamounts which are generally porphyritic and vesicular basalts with glassy to microcrystalline groundmasses, and which are predominantly alkali. Nayudu (1962) and Engel and Engel (1963) described olivine-rich to olivine-poor basalts from Cobb Seamount on the western flank of Juan de Fuca Ridge. The rocks include porphyritic, vesicular, and pillow basalts, and palagonite-tuff breccia. Two other seamounts from Eickelberg Ridge (Nayudu et al_, 1967) yielded porphyritic vesicular basalt, vesicular basalt, fine-grained basalt, pillow basalt, and palagonite-tuff breccias. Kay et_ al_ (1970) reported rock descriptions and chemical analyses of basalts from Juan de Fuca Ridge, Gorda Ridge, and Explorer Seamount. The basalts include subophitic, porphyritic, fine-grained, vesicular, and non-vesicular varieties; one sample from Gorda Ridge is doleritic. Most samples are low-potassium ocean ridge tholeiites with normative quartz or olivine. 20 The sample from Explorer Seamount contains normative nepheline. Kay et al found light-depleted rare earth element relative abundance patterns in almost all samples. Hedge and Peterman (1970) reported Sr87/Sr86 ratios ranging from 0.7012 to 0.7031 with an average value of 0.7026, in basalts from Juan de Fuca and Gorda Ridges and Explorer Seamount. They showed that these data combined with other data from the Pacific Ocean indicate that tholeiitic basalts being erupted along active ridges in the Pacific Ocean contain less radiogenic Sr^^ than basalts erupted on islands, such as the Hawaiian Islands. Studies of sediment cores from Cascadia Basin and Juan de Fuca Abyssal Plain west of the northern end of Juan de Fuca Ridge indicate major differences between Pleistocene and Holocene (since about 12,500 years B.P.) sedimentation (Griggs and Kulm, 1970; Carson and McManus, 1971). During the Holocene, sedimentation has been dominantly hemipelagic, resulting in the deposition of homogeneous olive-grey silty clays and clayey s i l t s . Turbidity currents have been restricted to channels, where olive-green s i l t sequences have been deposited. However, during the Pleistocene, extensive continental glaciations and lower sea levels resulted in voluminous sediment deposition on the outer continental shelves, and consequently, turbidity currents were much larger and more competent than during the Holocene. They deposited sand, s i l t , and clay not only in channels but on fans and abyssal plains. Hence, Pleistocene deposits underlying the hemipelagic Holocene sediments contain numerous layered sands and sandy silts (Carson and McManus, 1971). The sedimentation rate in the Cascadia Basin area was high during the Pleistocene. Duncan (1968) reported a late Pleistocene sedimentation rate of 43 cm/1000 years in eastern Cascadia Basin (Silver, 1971a). A 21 minimum rate for the Pleistocene turbidite section in the medial valley of Gorda Ridge, cored in the JOIDES Deep Sea Drilling Project, is 56 cm/1000 years (McManus et al_, 1970). Seismic reflection profiling (Ewing et al_, 1968; Hamilton, 1967) has enabled interpretation of turbidite sedimentation in the region. A thick wedge of sediment f i l l s Cascadia and Gorda Basins, and thins onto Juan de Fuca and Gorda Ridges, whose rough topography has acted as a barrier to sea-ward dispersal of terrigenous sediment from northern California, Oregon, Washington, and Vancouver Island. Hamilton and Menard (1968) suggested that uplift of Juan de Fuca Ridge after the beginning of turbidite sedimentation is indicated by the raised and starved flat-floored furrows on the ridge that contain turbidite deposits, McManus et al_ (1972) reached similar conclusions after a detailed seismic reflection survey at the northern end of Juan de Fuca Ridge. A s t i f f silty clay, Pliocene to Holocene in age, has been dredged from the lower continental slope east of Gorda Basin. It contains a lower bathyal fauna indicating a greater depth than that from which the sample was collected, suggesting late Cenozoic uplift of the area (Silver, 1971a). Similarly, sedimentary rocks dredged from the slope and shelf off central Oregon (Byrne et al_, 1966) and from the continental shelf west of Vancouver Island (Tiffin e_t al_, 1972) contain, fossil foraminifera of Pliocene to Miocene age indicating water depths much greater than those from which they were collected. A hole drilled by JOIDES on the lower continental slope west of Oregon encountered a turbidite section whose palaeobathymetric indicators suggest deposition in wafer 500 m deeper than the present depth (von Huene et al_, 1971). Continuous seismic reflection profiles across the shelf and slope from northern California to Vancouver Island show deformation 22 within the sedimentary sequence which has been interpreted to be the result of compressive forces (Byrne et a l , 1966; Chase et a l , 1970; Silver, 1969, 1971a; Tiffin et al , 1972). REGIONAL PLATE TECTONICS Tectonics of the northeastern Pacific Ocean are presently controlled by the relative motions of three lithospheric plates: the Pacific, North American, and Juan de Fuca Plates (Fig. 8)(Morgan, 1968). In the north, the boundary between the Pacific and North American Plates is the Queen Charlotte (Transform) Fault Zone (Chase and T i f f i n , 1972). The Triple Junction (Atwater, 1970) between the three plates in the north possibly occurs in the vicinity of the Dellwood Seamount area northwest of Vancouver Island (Bertrand, 1972). South of the Triple Junction, the boundary between the Pacific and Juan de Fuca Plates is composed of a series of spreading segments and transform faults—the Dellwood Spreading Segment, the Revere-Dellwood Fault, the Explorer Spreading Segment, the Sovanco Fracture Zone, Gorda Spreading Segment, and the Mendocino Fault. The southern Triple Junction is probably west of northern California where the Mendocino and San Andreas Faults intersect, the latter being a Pacific-North American Plate boundary (Silver, 1971b). Morgan (1968) pointed out that because the trends of the Blanco and Mendocino Fracture Zones differ from the trend of the San Andreas and Queen Charlotte Fault Zones, and also because of the many oblique faults shown on the Raff and Mason magnetic pattern as interpreted by Vine (1966), Juan de Fuca is an anomalous plate between the large Pacific and North American Plates, moving independently of the two larger plates. Silver (1969) used the rela-tive velocities and directions of motion of the Pacific, North American, and Juan de Fuca Plates to determine the rate and direction of differential 23 FIGURE 8. Boundaries of the Juan de Fuca Plate. 24 motion between the Juan de Fuca and North American Plates. Underthrusting of North America by Juan de Fuca is implied as well as oblique right-lateral slip along the continental margin (the inferred plate boundary). This situation appears to apply both for motions during the last 0.7 million years and for motions approximately 5 million years ago. Using the magnetic anomalies for the Northeastern Pacific Ocean and the tectonic features of western North America, Atwater (1970) discussed the Cenozoic history of the Northeastern Pacific. She showed that a trench probably existed west of North America throughout much of the Cenozoic, and that Juan de Fuca Plate is a remnant of the Farallon Plate which was consumed at that trench. Since about 29 million years ago the ridge i t s e l f , of which Juan de Fuca Ridge is a surviving remnant, has been progressively overrun by the trench. The details of the history of the region are equivocal. If the North American and Pacific Plates are assumed to have moved with a constant relative motion of 6 cm/year parallel to the San Andreas Fault then the Dellwood Triple Junction probably has been in the vicinity of Vancouver Island for at least 30 million years (Fig. 9, 10). Prior to that time, until about 80 million years ago, the Kula Plate lay off northwestern North America and the Kula-Farallon-North American Triple Junction was migrating towards the Aleutian Trench (Fig. 91, H, G). Strike-slip motion was progressively replaced by subduction as far north as Vancouver Island (Fig. 9E). Meaningful predictions concerning plate motions prior to 80 million years ago are impos-sible from the available data. However, i f a Mesozoic trench is required to produce the Franciscan rocks (Dietz, 1963; Hamilton, 1969; Ernst, 1970) and the Sierra Nevada and British Columbia Coast Mountains batholiths (Hamilton, 1969), then a major change in relative motions must have occurred about 80 million years ago. 25 A \ PRESENT F . 38 MY. 2250 KM B ' 5KT 300 KM FIGURE 9. Reconstruction of plate evolution and deformation related to Cenozoic interactions of the North American and Pacific plates and the North American and Kula plates. North America is arbitrarily held fixed and large arrows show notion relative to i t . Initials are cities listed in Figure 10. Diagrams in J show the derivation of various vectors. Captions give time in millions of years before present and amount of offset which must subsequently occur to bring the Pacific and inner North American plates back to their present relative positions. Pacific-North American motion (PA) is assumed constant at 6 cm/yr in a horizontal direction. For the last 20 my, 4 cm/yr is assumed to be taken up on near-coast faults while 2 cm/yr is accommodated by inland faults. Thus, California was moving northwest at 2 cm/yr (CA). Prior to 20 my ago, all motion is presumed to be taken up at the coast. Black regions are unacceptable overlaps of oceanic and continental crust, showing that the direction of Pacific-North American motion probably suffered at least a minor change sometime between 20 and 4 my ago. (After Atwater, 1971). 26 EPCIGS CH FIGURE 10. Locations of plate boundary regimes with respect to points in western North America assuming motions and deformations described in Figure 9. Inland cities are raised above the line. They have been projected to the coast vertically in the projection of Figure 9, roughly parallel to the direction of FaralIon-North American underthrusting. Fine lines trace the shifting locations of the cities as the continent deforms. White areas show times and places where North America was in contact with the Pacific plate or the Kula plate. The probable near coast manifestations of these interactions are stated. (After Atwater, 1971). 27 On the other hand, i f relative motions are not assumed to have been constant, then many different models are possible. Atwater (1970) dis-cussed one in which the North American and Pacific Plates were fixed to one another until 5 million years age, at which time they broke apart and began to move at a rate of 6 cm/year, parallel to the San Andreas. In this model, the Dellwood Triple Junction migrated down the coast of British Columbia during the past 30 million years, until about 5 million years ago when the Triple Junction reached the northwestern end of Vancouver Island (Fig. 11). Prior to about 30 million years ago a trench existed everywhere off western North America (Fig. HE) . In summary, i t is very probable that the continental margin west of Vancouver Island, Washington, and Oregon has been undergoing compressive stresses for at least 30 million years. OCEANIC BASALTS Many authors in the last decade have described rocks dredged from the ocean floors, or obtained from oceanic islands. As a result, the distri-bution of rock types in the ocean basins is reasonably well-known. By far the most common rock type is basalt. However, in spite of some experimental work on the origin of basalts (notably Yoder and Tilley, 1962; Green and Ringwood, 1967), no completely satisfactory model for oceanic basalt evolution has been proposed, although many, often conflicting, attempts exist in the literature. In this section, the more popular ideas on the subject of oceanic basalts are summarized. With these, the results from the study area may be compared (Chapter IV). 28 A . PRESENT B . 5HT 300 KN ,*s,,„t„,,y ->*^  7, G ION. I T r ^ & , V t , , , , - s . F L . f l NORTH GS . * •* 7 AMERICAN- . \ » |^V/Y PACIFIC \ 1 S/" PLATE V D. 20 M.Y. V 7 T r i T t r i i f i i i i i i . S F L * G S * V Z N O R T H A M E R I C A N ! P L A T E | ^5M^M / / > mV l , l , t f l l l V l, l l ;t F A R A L L O N PACIFIC PLATE OUEEN C H A R L O T T E S A N LOS I S L A N D S S E A T T L E F R A N C I S C O A N G E L E S G U A Y M A S M A Z A T L A N S T R I K E S L I P S T R I K E S L I P / . SUBOUCTBN *" AT TfiCNCM , N O R E L A T I V E - B E I N G T A K E N A T M A R G I N M O T I O N / U P - 1- - / - -SUSDUCTtON ; AT TRENCH" ' FIGURE 11, A to E. Schematic model of plate interactions assuming that the North American and Pacific plates were fixed to one another until S my ago, at which time they broke apart and began to move at a rate of 6 cm/hr. The coast is approximated as parallel to the San Andreas Fault. The Pacific plate is arbitrarily held fixed and large arrows show motion relative to it Captions show the time represented by each sketch in millions of years.before the present and the distance that the North American plate must subsequently be displaced to reach its present position with respect to the Pacific plate. (After Atwater, 1971). F. Evolution with time of boundary regimes along the coast of North America assuming the model of changing American-Pacific motions described in Figure 11. White areas denote times and locations of Pacific-North American interactions. Prior to 5 my ago these plates were locked, so that no plate-related tectonics should be detected in the regions and during the times marked "no relative motion . . .". White areas from 0 to 5 my ago indicate a regime of strike-slip. Grey areas show times and places where Farallon plate lay offshore and was underthrusting North America. Juan de Fuca plate is a remnant of the Farallon plate. (After Atwater, 1971). 29 Composition of Oceanic Basalts Engel et al 0-965) suggested that the dominant rock types in the ocean basins are what they termed the oceanic tholeiites. Subsequent workers have confirmed the predominance and rather consistent and unique chemical composition of these basalts (also called abyssal tholeiites, ocean ridge tholeiites) in the world's oceans (for example, Tatsumoto et a l , 1965; Muir and Tilley, 1966; Aumento, 1968; Frey et a l , 1968; Melson et a l , 1968; Miyashiro et a l , 1969; Kay et_ al_, 1970). In addition to these and other basalts of various types, altered basalts, metabasalts (greenschist and amphibolite), and coarser-grained rocks such as diabases, gabbros (and their metamorphic equivalents), diorites, serpentinites, and peridotites have also been dredged, particularly from fracture zones along the Mid-Atlantic and Indian Ocean Ridges (for example, Engel and Fisher, 1969). Throughout this thesis, the characteristic tholeiitic basalts, thought to compose at least 50% of the ocean crust (Engel and Engel, 1970), will be referred to as ocean ridge basalts (after Kay et a l , 1970). The characteristics of these basalts are: (a) High contents of Si02, AI2O3, and CaO compared to other basalts (b) Very low concentrations of K, T i , U, Th, Pb, Rb, Zr, Ba, and La compared to other basalts (c) Remarkably uniform rare earth element (REE) fractionation patterns, with patterns unfractionated relative to chondrites for heavy REE but depleted for light REE In general, ocean ridge basalts contain normative olivine, but some are slightly quartz-normative. High alumina and high-iron variants of ocean ridge basalts have been reported (Miyashiro et a l , 1969; Kay et a l , 1970). 30 These basalts are distinct from the other basalts of the oceans; (1) the alkali basalts (also called alkali olivine or alkaline) and (2) the Hawaiian and Icelandic tholeiites. The latter show generally higher Fe, T i , K, and Mg and less Al than the ocean ridge basalts (Engel, 1971) and different REE (light-enriched) patterns (Schilling, 1971). The alkali basalt, once thought to be the dominant rock type on the ocean floor is now considered to be largely confined to the suprastructures of oceanic volcanoes. These basalts contain higher abundances of the alkaline elements and U, Rb, Th, and Ba, and REE patterns enriched in the light REE and depleted in the heavy REE relative to ocean ridge basalts (Gast, 1970). Alkali basalts are nepheline-normative. Kay et al (1970, p. 1606), after studying the major and trace element contents of ocean ridge rocks from the Mid-Atlantic Ridge, Gorda Ridge, Juan de Fuca Ridge, and the East Pacific Rise, concluded that: ". . . i n general the characteristic olivine-tholeiite composition and dispersed-element composition . . . is in fact a distinguishing characteristic of rocks formed within the axial valley or axial zone of mid-ocean ridges. And, conversely, i t appears that seamounts, i.e., volcanism centered around an eruptive center, have more undersaturated and alkaline chemical affinities and are richer in larger cations." Origin of Oceanic Basalts As yet there is not a consensus of opinion regarding the interre-lations of the different basaltic magmas erupted at or near ocean ridges. Two main schools of thought have been presented: 1. Fractional crystallization under different pressure conditions of a single parental magma can produce most of the variations observed. Engel and Engel (1970) and Engel (1971) favoured this interpretation, suggesting that most, i f not a l l ; alkali basalts are derived from 31 tholeiitic parental magmas. They based their interpretation on occurrence, relative abundance, and compositional interrelation-ships of alkali and tholeiitic basalts. 2. Most petrologists stipulate at least two distinct parental magmas, one the alkali basalt, the other the tholeiitic basalt. Experi-mental petrologists have argued that the various basaltic magmas are formed by the partial melting of a source rock, probably of peridotite (or pyrolite) composition, under varying conditions of total pressure, water pressure, and temperature (Kuno, 1959, 1966; Yoder and Tilley, 1962; Nicholls, 1965; Green and Ringwood, 1967; Green, 1971). They suggest that the alkali magmas must be derived by lesser degrees of partial melting at greater depths in the mantle than tholeiitic basalts. This model explains the differences ^ in major element chemical composition. In addition, fractional crystallization processes may play an important role in producing the more minor variations observed among oceanic rocks. However, some elements (notably K, T i , Sr, Rb, U, Th, Pb, and the lighter REE) show variations which cannot be accounted for either by crystal fractionation or differences in degree of partial melting of the same parental peridotite. Their variations imply inhomogeniety in distribution of these elements in the source material. Aumento (1967) proposed a model for magmatic evolution on the Mid-Atlantic Ridge requiring multiple cycles of partial melting of the upper mantle related to tectonic patterns beneath the axis of the ridge to generate the different magma types (Fig, 12). The parental magmas are produced by different degrees of partial melting of a pyrolite (Green and Ringwood, 1967) mantle at a depth of 35 to 60 km, The most voluminous magmas produced are the olivine 32 W CREST MTNS. RECENT U E X T „ 0 S MEDIAN VALLEY 5 km. • ALKALI BASALT TRANS THOLEIITES LV.-: J OLIVINE THOLEIITES ppt. of olivine ppt. of olivine a low Co pyroxene ppt. of Mg-AI pyroxene % fractional melting sea level \- I km. 2km >- 3 km. — 15 km.-• 35 km-• 60km-< J = PYROLITE convection convection convection FIGURE 12. Model of magmatic evolution for the Mid-Atlantic Ridge (from Aumento, 1967). 33 tholeiites (with 14-16% AI2O3). From these as a result of fractional crystallization can be produced the high-alumina tholeiites (>16% AI2O3) and the quartz tholeiites. Lesser degrees of partial melting (due to lower mantle temperatures under the flanks of the ridge) yield transitional and alkali olivine basalts. Because the strontium isotopic compositions of these Mid-Atlantic Ridge samples show no variations between the olivine tholeiites from the Median Valley and the alkali basalts from the adjacent seamounts, a similar magma source for al l the rock types can be used. However this does not appear to be the usual situation (Hedge and Peterman, 1970), so the model is not generally applicable. However there does seem to be general agreement that high alumina tholeiite magmas at least are produced from the parent olivine tholeiite magmas by differentiation, particularly the crystal-lization of olivine. Kay et_ al_ (1970) suggested that the observed variations in the chemical compositions of the ocean ridge basalts (the sequence high-alumina basalt, olivine tholeiite, and iron-rich quartz tholeiite) can be explained by the low-pressure fractional crystallization of olivine and plagioclase. They concluded that the parent liquids of ocean ridge basalts are critically saturated (hypersthene- and olivine-normative) and have relatively high alumina contents (16-17.5%). Thus their parental magma contains considerably more alumina than that of Aumento (1967). Kay e_t al also concluded that the parent liquids have extremely low contents of large cations and chondritic rare earth element patterns. Hence i t is unlikely that alkali basalts or Hawaiian tholeiites could be derived from such a liquid. The major characteristics of this parental liquid are ascribed to extensive partial melting (30%) at shallow depths (15-25 km) in mantle material being intruded under the axial zones of mid^ocean ridges. Gast (1970) suggested that, i f the magma then ascends 34 rapidly with negligible fractional crystallization, the typical ocean ridge basalt is extruded at the surface. Cast also suggested that basaltic liquids forming oceanic islands and searaounts are formed at much greater depths within a partially (3%) molten low-velocity zone in the mantle. These liquids may then undergo fractional crystallization at shallow depths to produce variants. Bass (1971) suggested a systematic variation of basalt chemistry as a function of spreading rate. More slowly spreading ridges tap magmas from greater depths (>25 km) which reach the surface without long stays in the 0-5 kbar pressure region, and are richer in alumina and more undersatu-rated, compared to faster spreading ridges. The latter tap magmas from shallow depths (25-15 km) which are low in alumina and tend to be quartz-and strongly hypersthene-norrnative. They apparently undergo extensive low-pressure plagioclase-olivine fractionation to produce high-alumina, high-iron basalts. Magmas formed at very great depths are alkali basalts. Aumento's (1967) model for the Mid-Atlantic Ridge at 45°N can be fitted into Bass' idea because Aumento derived the low-alumina olivine tholeiite from much deeper (35-60 km) than Bass visualized for a typical slow-spreading ridge. The area studied by Aumento (the Mid-Atlantic Ridge at 45°N) would thus appear to be anomalous, even compared to other areas on the Mid-Atlantic Ridge. That a fundamental difference may exist between ocean ridge basalts from slow- and fast-spreading ridges, or at least between Atlantic and Pacific ocean ridge basalts, is further supported by strontium isotopic data, which suggests that Mid-Atlantic Ridge ocean ridge basalts are more radiogenic than those from the Pacific (Tatsumoto et a l , 1965; Hedge and Peterman, 1970). 35 Hence, i t appears that some variations observed in the compositions of the primary ocean ridge basalt can be related to depths of origin of the parent magmas, whereas other differences can be related to varying degrees of fractional crystallization. CHAPTER III STRUCTURAL FEATURES OF THE OCEAN FLOOR IN THE STUDY AREA BATHYMETRY Bathymetric Chart Preparation Because considerable bathymetric information was obtained during cruises to the study area, a revised bathymetric chart was prepared in two sections: 1. The northern end of Juan de Fuca Ridge, the Sovanco Fracture Zone, and the seamounts in the area bounded by 47.9°N and 49.1°N latitudes, between 128.0°W and 129.5°W longitudes. The data available from this area included: (a) Bathymetric data collected during the survey in the region by the U.S.C.G.S. ship PIONEER (This data was supplied by J. Mammerickx Winterer, Scripps Institue of Oceanography). A radio-navigation system, with fixed beacons ashore, was used in this survey, and enabled an accuracy in positions of about 0.1 mile. (b) Bathymetric data collected by the Department of Oceanography, University of Washington, during a 1970 cruise to the area (made available by D. A. McManus, University of Washington). Positioning was by satellite navigation. (c) Depths obtained from echo sounder records made during the author's cruises to the study area in 1969, 1970, and 1971. 37 Positioning was generally by Loran A, augmented by satellite navigation on some profiles. Ship tracks from which bathymetric data were obtained are shown in Figure 13, 2. The continental slope, between 47.9°N and 49.1°N latitudes. The data available from this area included the PIONEER survey data and data collected during this project, as described for the ridge crest area. In addition, some depths were obtained from profiles located by satellite navigation in the southern part of the area, recorded during a cruise of the United States Geological Survey (copies of these profiles were supplied by W. Barnard, University of Washington). Ship tracks from which bathymetric data were obtained are shown in Figure 14. Because of the paucity of any data additional to the PIONEER survey data, and, in Cascadia Basin, lack of relief, the remainder of the study area was not recontoured. On the revised bathymetric chart (Fig. 15, p. 287) contours for the non-revised areas are taken from Mammerickx and Taylor (1971). In a l l cases the original data are in fathoms. In area 1 (between 128°W and 129.5°W) a computer program was used to convert depths to metres, correct them for the variable velocity of sound in water according to tables of Matthews (1939), plot them, and produce a contour chart. To prepare the final version, a computer plot of spot-depths was contoured by the writer, using the computer's mathematically correct interpretation as a guide, but modifying i t where another interpretation seemed more reasonable. Because of the amount of modification necessary (usually due to small navigational errors 38 FIGURE 13. Ship tracks from which were obtained bathymetric data used in preparation of revised bathymetric chart of northern end of Juan de Fuca Ridge (Fig. 15, p. 287). Data from U.S.C.G.S. ship PIONEER survey, University of Washington surveys, and the present investigation. Base map is from Figure 15. 39 FIGURE 14. Ship tracks from which were obtained bathymetric data used in preparation of revised bathymetric chart, continental slope area (Fig. 15, p. 287). Data from U.S.C.G.S. ship PIONEER survey, U.S.G.S. seismic re-flection survey, and the present investigation. Base map is from Figure 15. 40 and discrepancies between the different sets of data) i t was decided not worthwhile to contour the slope data by computer, and i t was done only by hand. The revised bathymetric chart for the study area is presented in Figure 15 (p. 287). This is used as a base map for data within the study area presented in this thesis. However, when data from a larger area than the study area are being presented, then the base map is from Mammerickx and Taylor (1971). Physiography On the basis of physiography (Fig. 15, p. 287) the study area is divided into six regions—the continental slope, Cascadia Basin, Juan de Fuca Ridge, Sovanco Fracture Zone, the seamount chains, and the abyssal plains. 1. Continental Slope The continental slope includes the area between the continental shelf break (at a depth of about 200 m) and Cascadia Basin (depth about 2500 m at the base of the slope). It has an average width of 70 km. The trend of the continental slope is northwest. The lower continental slope is characterized by discontinuous ridges and troughs, which generally parallel the trend of the slope. In one place the first ridge rises as much as 800 m above the adjacent Cascadia Basin. The middle and upper slope rise rather uniformly to the shelf break, but the topography is considerably modified by several generally southwest- or west-trending submarine canyons. From north to south, the most prominent canyons are Clayoquot, Father Charles, Loudoun, Barkley, Nitinat, and Juan de Fuca. In the southern part of the area, the base of the continental slope and eastern Cascadia Basin are modified by Nitinat Fan, built out into Cascadia Basin by coalescing of fans below Barkley, Nitinat, and Juan de Fuca Canyons. 41 Between 48°50'N and 49°N latitudes the abrupt rise from Cascadia Basin is followed by a wide plateau, at a depth of about 2050 m, northeast of which the typical middle and upper slope features are present. 2. Cascadia Basin Cascadia Basin lies between the continental slope and the higher topography of Juan de Fuca Ridge. It also includes the smooth-floored area north of Juan de Fuca and Sovanco Ridges. The sea floor in Cascadia Basin is a grossly featureless depositional surface grading south with a slope of 1:850. In the southeastern part of the study area Nitinat Fan has been built out into Cascadia Basin. North of Juan de Fuca Ridge and Sovanco Ridge Revere Channel crosses the basin towards the southwest (Fig. 1). It apparently crosses Sovanco Fracture Zone near its junction with Explorer Ridge, perhaps enabling turbidity currents carrying terrigenous sediment to reach the abyssal plains on the western flank of Juan de Fuca Ridge. East of Juan de Fuca Ridge, Cascadia Basin is traversed by Vancouver Channel. Although this feature is broad and poorly developed in the study area, i t narrows and deepens to the south, where it eventually joins Cascadia Channel west of Oregon. 3. Juan de Fuca Ridge North of about 48°20', Juan de Fuca Ridge consists of three sub-parallel ridges, trending N 10°E. In this thesis these ridges will be referred to as East, Middle, and West Ridges, and the intervening valleys as East, Middle, and West Valleys, the latter lying west of West Ridge (Fig. 15). The relief associated with these features is variable. 42 East Ridge forms the western margin of Cascadia Basin, rising only about 100 m above the level of the basin. East Valley is the same depth as Cascadia Basin, and merges with i t north of East Ridge. Middle Ridge is most prominent near its northern end, where i t is as shallow as 1850 m, and only 10 km wide. To the south, Middle Ridge broadens and its relief diminishes. Middle Valley does not extend as far south as Middle and West Ridges. Between the ridges i t is about 14 km in width, but widens to 22 km north of West Ridge where i t lies between Middle Ridge and the high topography of Sovanco Ridge. Farther north i t merges with Cascadia Basin. West Ridge and West Valley are the most prominent of the series. Evidence is marshalled in Chapter V to indicate that West Ridge and West Valley are the youngest features of Juan de Fuca Ridge and the site of most recent sea floor spreading. The crest of West Ridge is less than 2200 m deep for much of its extent, rising over 700 m above the floor of West Valley. At its northern end West Ridge merges with the high topography of Sovanco Fracture Zone, to terminate West Valley. To the south, West Ridge terminates abruptly at Endeavour Trough*, and West Valley is cut off by Endeavour Seamount and the easternmost peak of the Heck Seamount Chain. Endeavour Seamount does not appear to be part of the Heck Seamount Chain, being displaced to the northeast relative to the trend of that Chain. South of Middle Valley and Middle Ridge the sea floor is smooth and slopes gently eastward to merge with East Valley, or, south of East Ridge, with Cascadia Basin. l7he names Endeavour Trough and Endeavour Seamount are suggested for these previously unnamed features by the author after the C.F.A.V. ENDEAVOUR. 43 4. Sovanco Fracture Zone The topographic expression of Sovanco Fracture Zone is a broad, northwest-trending ridge, over 100 km in length, with hummocky crestal topography, hereafter referred to as Sovanco Ridge. Minimum depth on Sovanco Ridge is about 2000 m, although much of the crest is over 2300 m deep. Although Sovanco rises over 500 m from the plain to the south, i t is in places only 100 m above the level of the sea floor in Cascadia Basin to the north. Eastward, Sovanco Ridge terminates abruptly at Middle Valley, but seems to merge with West Ridge to the southeast. Westward, Sovanco broadens and merges into topography of high relief on the eastern flank of Explorer Ridge. 5. Seamount Chains Two northwest-trending seamount chains l i e within the study area. Heck Seamount Chain consists of three distinct seamounts which coalesce at their bases. The largest and highest is Heck Seamount it s e l f , whose summit lies 1080 m deep. South of the Heck Chain, the Heckle Chain has only one distinct peak, at its western end, which has a summit at a depth of 1339 m. Both chains are about 13 km in width; Heck Chain is 70 km in length, Heckle 80. Eastward, these seamount chains merge into the crestal topography of Juan de Fuca Ridge, and westward a strip of abyssal plain 10 km wide separates them from the hilly topography on the eastern side of Explorer Trough. To the north and south, both seamount chains are flanked by abyssal plains. Some confusion exists in the literature regarding the identity of Heck Seamount. A seamount in the region was originally named after Captain Nicholas Heck of the U. S. Coast and Geodetic Survey in the 1940's prior to 44 the existence of detailed and accurately located bathymetric data (McManus, 1964), On his bathymetric chart 0-964) McManus incorrectly identified as Heck Seamount a large seamount northeast of the seamount originally named "Heck" (McManus, pers. comm., 1970). McManus et al (1972) revert to the original terminology. However, because the seamount to the northeast is the most prominent in the region, and because the recent detailed and accurate bathy-metric chart of the region identifies that seamount as Heck (Mammerickx and Taylor, 1971), that seamount is called Heck in this thesis, and the chain in which i t lies is termed the Heck Seamount Chain. It is suggested here that the seamount and seamount chain to the southwest be referred to as Heckle. 6. Abyssal Plains Smooth, flat sea floor averaging 2800 m in depth flanks the seamount chains. McManus et^  al_ (1972) referred to this area as Juan de Fuca Abyssal Plain. Between the eastern ends of Heck and Heckle Seamount Chains the plain shallows abruptly to an irregular level of about 2600 m, extending to the low ridge which trends south from the eastern end of the Heck Seamount Chain and which apparently marks the western margin of the crestal area of Juan de Fuca Ridge. A low h i l l (whose shape and extent are poorly defined by available data) lies on this shallow level between the seamount chains. Similarly, between Heck Seamount Chain and Sovanco Fracture Zone the abyssal plain shallows abruptly to about 2700 m and there the shallow region is complicated by a northwest-trending ridge which protrudes into the abyssal plain. East of this shallow region is West Valley. These irregular, shallow areas may represent poorly developed seamounts or seamount chains. However, the shallow region west of West yalley between the eastern end of Heck Seamount Chain and Sovanco Fracture Zone may also have formed either as a consequence of regional uplift or as a direct result of sea floor spreading (see p. 180). 45 SEISMIC REFLECTION DATA Introduction Approximately 2200 km of continuous seismic, reflection profiles were recorded during the seven cruises to the study area, 1000 km over the continental slope, 500 km over Juan de Fuca Ridge, and the remaining 700 km from the other regions within the study area. The quality of the records varies because several different combinations of sound source and recorder were used to obtain them. Records of best quality, including most reproduced here, were obtained using a Bolt airgun (Model 600B) with a 10 cubic inch chamber, firing at 2000 psi, and a wet-paper Gifft Graphic Recorder, during cruise 70-16 on the C.E.A.V. ENDEAVOUR. On other cruises a 5000-joule sparker system (described by T i f f i n , 1970) and a wet-paper Alden Precision Graphic Recorder or a dry-paper EPC recorder were used. For all profiles the returning echoes were received by a linear array of 20 hydrophones (Geospace M-7), which were towed about 200 ft behind the ship. This information was augmented by copies of 850 km of profiles mostly from the north-central part of the study area obtained during cruises of the Department of Oceanography, University of Washington, under the direction of D. A. McManus (who kindly supplied the copies). Subsequently most of these profiles were published (McManus e_t a l , 1972). A representative collection of the seismic profiles is reproduced in this thesis (Plates I to XIX). Magnetic data were recorded simultaneously with many seismic profiles, and residual magnetic anomaly profiles are plotted above those profiles. The residual anomaly profiles were calculated by sub-tracting from the original magnetic data the intensity of the geomagnetic reference field (see p. 84). Line-drawing interpretations only are reproduced of two profiles which are important but of poor quality, together with one magnetic profile (Fig. 23). 46 The locations of the seismic and magnetic profiles in Plates I to XIX and Figure 23 are plotted in Figures 51 and 52 (Appendix A). The methods and difficulties of geological interpretation of seismic reflection profiles have been thoroughly discussed by many authors (discus-sions by Moore (1968), Tiffin (1970), and Van Andel and Heath (1970) have particular application to this study) and hence, will not be discussed in detail here. The major interpretive difficulty encountered in the profiles used in this study was the identification of side echoes. In places, this problem could be resolved by studying other profiles which cross the same or similar features at different locations and angles. Numerous faults are inferred on the line-drawing interpretations. Because steep slopes (those exceeding about 1 5°; Moore, 1968) are not generally recorded by the seismic profiling technique, there is no direct evidence of fault planes on the profiles; their presence is deduced from criteria such as those discussed by Moore (1968) and Van Andel and Heath (1970). The faults drawn on the interpretations do not necessarily portray the actual fault-plane attitudes, but indicate only that faults are interpreted to be present. All the reflection profiles and their interpretations shown in Plates I to XIX are exaggerated in the vertical scale. The degree of vertical exaggeration, i.e., the ratio of horizontal to vertical scales, is indicated on each profile. In addition, representative nomographs (Fig. 53) show the true sea floor slopes. Vertical exaggeration and nomographs were calculated using a sound velocity of 1.5 km/second. Due to changes in ship speed, hori-zontal scale and hence vertical exaggeration in some cases vary along one profile. A key to the symbols used in the line-drawing interpretations is given in Figure 53. 47 Geologic Units Juan de Fuca Ridge Area On the basis of the seismic profiles, a geological map of the Juan de Fuca Ridge area was compiled (Fig. 16), The units shown on this map are described below. 1. Igneous Basement The igneous basement is evident on a l l the seismic profiles except those from the continental slope and from parts of Middle Valley where the sediment thickness exceeds the penetrative ability of the seismic signal. The interpretation of this basement as igneous rock is based upon two lines of evidence: (a) The characteristic nature of its reflections, which are relatively strong, with an abundance of hyperbolae from point reflectors (indicating a rough and irregular topography), and the lack of sound energy penetration into the basement. (b) The dredging of basalt from 13 outcrops of this basement selected from the seismic profiles on Juan de Fuca Ridge, Sovanco Ridge, and the seamount chains. Because other types of igneous rocks (or their metaraorphic equivalents) have been dredged from the ocean floors in other regions, although but rarely compared to basalts, i t is remotely possible that the igneous basement in the study area is not every-where basaltic. A chart of the depth of igneous basement below sea level (Fig. 17) demonstrates that the relief of the basement surface is much greater than the topographic relief in most places, especially on the crestal region of Juan de Fuca Ridge, For example, the basement relief along Middle Valley in places FIGURE 16. Geological map of the Juan de Fuca Ridge area north of 48°N latitude. Faults are drawn perpendicular to the seismic reflection profiles on which they are observed. Only major faults are shown. Geologic units are described in the text. FIGURE 17. Depth below sea level to igneous basement in seconds of 1-way travel time. Legend indicates approximate metre equivalents assuming a sound velocity of 1500 m/sec in water and sediment, which provides a minimum depth. The deepest basement is below Middle Valley. Areas more shallow than 1.4 seconds are not contoured. 50 exceeds 1 second (one-way travel time), or probably at least 2000 m. As indicated on the geological map (Fig- 16), this relief is apparently largely due to faulting. Where i t is not exposed on the sea floor, the igneous basement is covered by a varying thickness of sediment (Fig. 18). 2. Sedimentary Sequence The reflectors generally observed within the sedimentary sequence overlying the igneous basement in seismic profiles from the study area are characteristic of turbidite sequences. They are attributed to changes in acoustic impedance (i.e., changes in the product of density and compressional velocity of the respective layers), probably related to changes in grain size, lithology, or degree of compaction. However in some parts of the sedimentary sequence, reflectors are weak or absent. In places, this could be due to a general decrease in relative and absolute contrasts in impedance or increased absorption of the signal spectrum related to increasing depth of burial, or to homogeneity of the sediment that was deposited. However, particularly on the continental slope, there is evidence that deformation is an important factor in destroying internal reflectors. Pelagic and hemipelagic sedimentation (which can also produce sedi-mentary sequences lacking strong internal reflectors) are thought to have played only minor roles in the study area because of the probable young age of the igneous basement (Fig. 4), the large volumes of sediment that have been deposited, and the comparatively slow rates of pelagic sedimentation. On the western flank of Juan de Fuca Ridge south of the study area where, because the ridge acts as a barrier to sediment dispersal by turbidity currents, sedimen-tation is almost certainly pelagic or hemipelagic, only 2 m of sediment are FIGURE 18. Minimum sediment thickness over the Juan de Fuca Ridge area north of 48°N latitude. Thicknesses were calculated assuming a velocity of sound in sediment of 1500 m/sec. Maximum sediment thickness is in Middle Valley. 52 reported (Lister, 1970). Within the study area the seamounts and parts of Juan de Fuca and Sovanco Ridges are virtually devoid of sediment cover (Fig. 18). If either pelagic or hemipelagic sedimentation were exceptionally rapid in the region, then these areas would be expected to have sediment cover as well. Consequently, the main sources of sediment in the study area are considered to be turbidity currents, whose loci are controlled by the topography of the sea floor. The thickness of sediment in the.study area (Fig. 18) was calculated using a sound velocity of 1500 m/second. This value gives a minimum sediment thickness. Moore (1968), on the basis of velocity measurements, found this velocity to apply for sediment thicknesses to 0.25 seconds (one-way travel time). For thicknesses between 0.25 and 0.50 seconds, Moore used a velocity of 1900 m/second, and for thicknesses greater than 0.50 seconds, a velocity of 2300 m/second. However, because sediment velocities can be very variable (for example, McManus et a l , 1972) and no velocity measurements are available for the area, and because the sediment thickness is less than 0.25 seconds over much of the Juan de Fuca Ridge area, the thicknesses were calculated using 1500 m/second and are considered to represent minimum thicknesses. Sediment thickness (Fig. 18) varies over the study area from zero (confirmed by dredging at some locations in the area), to over 1600 m in Middle Valley and 2000 m in eastern Cascadia Basin near the base of the con-tinental slope (assuming average sound velocities of 2000 m/second, probably reasonable for such thick sections). The sediment thickness below the continental slope is not known from seismic profiling. Three sedimentary units, whose outcrop distributions are indicated on Figure 16, are recognized in the Juan de Fuca Ridge area: 53 a. Unit A This sequence of turbidites directly overlies igneous basement in much of the study area, forming the ocean floor on many topographic "highs", including East Ridge and parts of Middle Ridge, West Ridge, and Sovanco Ridge, and probably in West Valley (Fig. 16). This unit corresponds to the Lower Turbidites of McManus ejt a_l_ (1972) , and hence may consist of two sub-units, a lower, more acoustically transparent unit and a thin upper unit containing strong reflectors. This division is clear in profiles from Cascadia Basin north of the study area (McManus et a l , 1972). However, the subdivision of Unit A within the study area is not possible. The unifying characteristic of the unit is the intense deformation within i t , mainly faulting, but also tilting and folding. Unit A has been deformed together with the underlying igneous basement, but the basement ini t i a l l y had some relief which controlled Unit A deposition. Because the unit is more widespread than Unit B, the relief of the sea floor during Unit A deposition was apparently much less than at present. The appearance of Unit A and its relationships to the igneous basement and Unit B are illustrated in Figures 20 and 21, which are line-drawing interpretations of seismic profiles from both Appendix A of this thesis and McManus et al_ (1972), whose locations are shown in Figure 19. These profiles also illustrate how Unit A can be traced over much of the Juan de Fuca Ridge area with l i t t l e ambiguity because of profile intersections. This is further demonstrated by a fence diagram of the profiles of Figures 20 and 21 drawn in isometric projection (Fig. 22). It is not known what the Unit A-Unit B contact represents. It appears that much of the region may have undergone synchronous, abruptly initiated deformation at the end of Unit A time, as suggested by McManus et al 54 FIGURE 19. Locations of seismic profiles illustrated in Figures 20 and 21, and the fence diagram in Figure 22. Bathymetry is from Figure 15. Capital letters designate profiles from this study, reproduced in Appendix A. Small letters refer to profiles from McManus et a l , 1972. FIGURE 20. Line-drawing interpretations of seismic reflection profiles to illustrate the distributions and interrelationships of sedimentary units A and B. Locations of profiles are shown in Figure 19. Capital letters designate profiles from this study reproduced in Appendix A. Small letters refer to interpretations of profiles of McManus et a l , 1972. Arrows indicate crossings of profiles. Igneous basement is patterned. FIGURE 21. Line-drawing interpretations of seismic reflection profiles to illustrate the distributions and interrelationships of sedimentary units A and B. Locations of profiles are shown in Figure 19. Symbols as in Figure 20. X o o X Unit B Unit A Igneous Basement V.E. X 6.25 FIGURE 22. Fence diagram (isometric projection) of interpretations of seismic reflection profiles illustrated in Figures 20 and 21 to show structure and distribution of geologic units near the northern end of Juan de Fuca Ridge. Locations of profiles are shown in Figure 19. Minor faults are omitted and profiles simplified for clarity.: 58 (1972). Following this event, minor deformation continued, but the nature of subsequent sedimentation (Unit B) was different. Alternatively, the process may be continuous, and part of Unit B may be synchronous with parts of Unit A, Due to the difficulties of sediment distribution over an area of appreciable relief, this alternative seems unlikely. A unique appearance of the northern end of Juan de Fuca Ridge at the present time is apparently required. b. Unit B Generally, Unit B unconformably overlies Unit A, and forms the uppermost sequence on the seismic profiles. This unit is restricted to the lowlands--the valleys on Juan de Fuca Ridge, the abyssal plains, and Cascadia Basin (Fig. 16). Unit B generally contains strong, essentially horizontal reflectors. Only minor deformation has affected this unit, including slight warping and tilting in the lower parts, and some minor faulting throughout. c. Unit C Unit C is restricted in distribution to regions above the abyssal plain level. The sediment composing this unit is considered to be locally derived from adjacent basement highs, particularly seamounts. On the abyssal plains near seamounts, this same material probably forms a large component of Unit B. Continental Slope On the continental slope, three sedimentary units are also recog-nized on the seismic profiles. These units probably overlap in age and hence are time transgressive stratigraphic units, because of the processes inferred 59 for the formation of the continental slope in the area (to be discussed in detail in the next section on structural interpretation). At this point, only the identifying characteristics of these three units are discussed: 1. Sedimentary "Basement" This unit behaves as acoustic basement on the seismic profiles, but it is not igneous basement because: (a) In three seismic profiles, the igneous basement can be seen below the sedimentary basement (b) Its reflectors lack the characteristic strong and irregular appearance of igneous basement. (c) Mudstone and siltstone have been dredged from outcrops of the sedimentary basement. However, the unit is acoustically reflective and is characterized by a nearly complete absence of internal reflectors. Its upper surface is irregular, and appears to be controlled by block faulting. The unit is exposed in some places on the lower continental slope. 2. Stratified Continental Slope Unit This unit overlies the acoustic basement, generally draping its irregular surface but lying flat in basement depressions, on plateaus, and in canyons. It invariably contains strong reflectors. Lower parts of the unit are in places warped, or faulted. 3. Pleistocene Wedge This unit unconformably overlies stratified material at the shelf break, forming a wedge-shaped cap. Microfauna in dredged samples suggest a Pleistocene age (B.E.B. Cameron, pers. comm., 1972) for this unit, whose 60 deposition may have been during a low stand of sea level, when parts of the continental shelf were subaerial (Carter, 1971). Probably this unit is con-temporaneous with part of the stratified slope unit farther down the continental slope. Structural Interpretations For each region previously defined on the basis of physiography, the structural elements are discussed below. When seismic reflection profiles are alluded to, capital letters refer to profiles recorded as part of this study and reproduced as plates in Appendix A. Non-capital letters refer to profiles published by McManus et_ al_ (1972) whose line-drawing interpretations are in Figures 20 and 21. Locations of the former profiles are shown on Figures 51 and 52, Appendix A; locations of the latter on Figure 19. 1. Juan de Fuca Ridge The physiography of Juan de Fuca Ridge (Fig. 15, p. 287) reflects the topography of the igneous basement, but the latter has been subdued by transport of sediment into the valleys to create a comparatively smooth sea floor. The basement relief (Fig. 17) in some places exceeds one second (in the order of 2000 m), whereas the maximum topographic relief of the sea floor is only about 800 m. The seismic profiles indicate that this basement relief is controlled mainly by faulting (Fig. 16). Many of the faults shown on Figure 16 may be continuous, or they may form series of en echelon faults such as those along the Central Gregory Rift in Kenya, East Africa (King, 1970; B. C. King, oral comm., 1972). Neither the number of seismic profiles nor the accuracy of their locations is adequate to allow one to trace with certainty most faults from profile to profile. 61 In the following discussion, a velocity of sound in sediment of 1500 m/second is assumed for the calculation of basement relief--in most places this probably results in an understatement of relief. The seismic reflection data indicate that the in i t i a l configuration of Juan de Fuca Ridge in the study area (revealed by the distribution of Unit A turbidites) was similar to but of much less relief than the present configuration. The secondary relief was apparently developed rather abruptly during a period of deformation which, probably fortuitously, occurred approximately at the end of Unit A time. East Ridge East Ridge appears to be a continuous basement feature extending from about 48°-48°45'N latitude. It consists of one (near its northern end) or more eastward-tilted fault blocks, with a basement scarp forming the western flank (Profiles B-B», C-C, D-D', E-E'--Plates II, III, IV, V; Profile f-f'--Fig. 21). Unit A overlies the igneous basement on East Ridge and has been faulted and tipped along with the basement. The Unit B sequence onlaps the eastern flank of the ridge and nearly f i l l s East Valley to the west, and shows only minor warping, t i l t i n g , and faulting. East Ridge is of maximum relief about 20 km south of its northern end, where its crest rises 180 m above the adjacent sea floor (Profile f - f1- -Fig. 21). It flattens, broadens, and dies out gradually toward its southern end (Profiles D-D', C-C), and near 48° is only an irregular basement high (Profile B-B1). The relief on the basement fault scarp forming the western flank decreases from 900 m in Profile f-f* to 550, 375, and 50 m in Profiles D-D', C-C», and B-B'. Toward its northern end, East Ridge also flattens and the relief on the western fault scarp decreases (Profile E-E'). There is no bathymetric 62 evidence for the continuation of the ridge north of the latitude of Profile E-E', and i t does not appear even as a basement feature in Profile e-e' (Fig. 21). Hence East Ridge apparently terminates abruptly at about 48°45'. East Valley East Valley, a fault-controlled valley between East and Middle Ridges, apparently terminates at the same latitude as East Ridge in the north. From there south to about 48°10' i t is a well-developed feature in every seismic profile (Profiles E-E', D-D1, C-C'--Plates V, IV, III; Profile f-f'--Fig. 21). It is bounded on the east by the fault scarp already described as the western margin of East Ridge; on the west, by a smaller fault scarp and the sloping flank of Middle Ridge. A thick sequence of Unit A f i l l s the bottom of the valley and thins over the western flank. Within the valley the contact between Unit A and overlying Unit B is observed at a depth of 2670 m in a l l profiles, although, on the western flank, Unit A strata are tilted and uplifted to shallower depths. The thickness of Unit A in East Valley decreases from 625 m near its northern end to 275 m in the south. On the basis of these observations, i t is suggested that East Valley was a topographic depression whose floor had a slope downward to the north during deposition of Unit A. The depression f i l l e d and overflowed onto adjacent East and Middle Ridges; then deformation resulted in uplift and tilting of the strata of Unit A on the ridge prior to deposition of Unit B. East Valley is not observed in the most southerly profiles (A-A', B-B'--Plates I, II), in which the igneous basement west of East Ridge forms a broad depression f i l l e d with horizontal strata of Unit A and Unit B. 63 Middle Ridge Middle Ridge, like East Ridge and East Valley, is most prominent near its northern end where its crest is as shallow as 1850 m (Profile F-F'--Plate VI). Like East Ridge, Middle Ridge consists of one eastward-tilted fault block near its northern end, and the number of fault blocks increases as the ridge deepens and broadens to the south (Profiles E-E', D-D', C-C--Plates V, IV, III; Profile f-f*--Fig. 21). Near its southern end, Middle Ridge consists only of an irregular, faulted basement high (Profiles B-B', A-A'--Plates II, I). Unit A overlies igneous basement over the broader southern part of Middle Ridge, and has been faulted and tilted with the basement (Profiles f - f , D-D', C-C, B-B', A-A'). Towards the northern end, Unit A overlies only the eastern flank of the ridge (Profiles E-E', F-F'). On the basis of Profile F-F', the following sequence of events is postulated in that region: Unit A was in i t i a l l y deposited on a basement surface dipping approximately 8° to the east; basement and Unit A were subsequently uplifted and tilted an additional 8° to the east prior to deposition of Unit B. Minor tilting and faulting within Unit B suggests that this deformation has persisted in moderated form to the present. The sediments on the western flank of Middle Ridge (Profiles F-F1, E-E') are interpreted to be part of Unit B on the basis of comparison with Profile f - f (Fig. 21) and Profile 41, Figure 7, in McManus et jd (1972), which clearly show that these sediments are faulted equivalents of part of Unit B in Middle Valley. North from Profile F-F", the relief of Middle Ridge decreases, and an increasing thickness of Unit A sediments covers the ridge (Profile G-C--Plate VI). The igneous basement dips to the north at about 2 . 8° , and 64 reflectors within overlying Unit A have approximately the same dip, more evidence of basement uplift and tilting after deposition of Unit A and prior to deposition of Unit B. The on-lapping Unit B strata are essen-tially horizontal, although minor tilting and faulting affect that unit as well (Profile c-c'--Fig. 20). Faults on Profile G-G' may be subsidiary faults related to the western boundary fault of Middle Ridge, or they may be independent trans-verse faults related to the northern termination of Juan de Fuca Ridge. Middle Valley Middle Valley is controlled by major faults on its northern, eastern, and western margins, but a seismic profile over the southern margin (Profile g-g'--Fig. 21) shows no identifiable faults. Instead, the igneous basement and overlying Unit A strata appear to be downwarped to form the valley. In contrast, a profile across the northern margin (Profile b-b1--Fig. 20) shows that both basement and Unit A are down-faulted to form Middle Valley. Hence, like East Valley, Middle Valley was largely formed after deposition of Unit A and prior to deposition of Unit B, although the thickening of Unit A in Profile g-g* (Fig. 21) suggests that a sea floor depression did exist during deposition of that unit. Minor deformation within Unit B (Profile b-b'--Fig. 20) again indicates that movements have continued in moderated form up to the present. The eastern margin of Middle Valley is the faulted western flank of Middle Ridge (Profiles H-G, F-F", E-E', D-D'—Plates VII, VI, V, IV; Profiles d-c', e-e', f-f'--Fig. 20, 21). 65 In the wider northern part, the western margin of Middle Valley is a steep fault scarp forming the eastern face of Sovanco Ridge (Profiles H-G, d-c'). The vertical displacement of basement on this fault is apparently at least 2000 m. Further south where Sovanco Ridge merges with West Ridge, a basement high associated with West Ridge protrudes into the western part of Middle Valley (Profile e-e'--Fig. 21). In this profile, Unit A can be traced from Sovanco Ridge into the western part of the valley. Where West Ridge forms the western margin of Middle Valley, the boundary is apparently a fault (Profiles E-E1, D-D', f - f ) . The sedimentary section thins towards the southern end of the valley, and in Profile D-D', the basement can be identified below at least 700 m of sediments. In general, the contact between Unit A and Unit B cannot be identi-fied within Middle Valley. Unit B is more deformed than is typical, particu-larly in the eastern part of the valley on the flank of Middle Ridge. Interestingly, those profiles which cross Middle Valley show faulting of Unit B, whereas the longitudinal profile b-b' (Fig. 20) shows only gentle warping. This indicates that the faults approximately parallel the trend of the valley, and i t is suggested that they are caused by east-west tension in Middle Valley, accompanied by uplift of Middle Ridge. West Ridge Unlike Middle and East Ridges, West Ridge does not die out to the south but is cut off abruptly, perhaps at a transverse fault zone. The eastern margin of West Ridge is fault-controlled, except south of Middle Valley where tilted Unit A strata lie on the gently dipping eastern flank, indicating uplift of the ridge after their deposition (Profile C-C--Plate III). The western 66 margin of the ridge appears to be only partly fault controlled (Profiles E-E', D-D', C-C'--Plates V, IV, III; Profile f-f'--Fig. 21). No sediments overlie basement on West Ridge, indicating either that i t has always been high, or that recent extrusion of magma has occurred there. Such extrusion might also have obscured faulting on the western flank. In the northeast, West Ridge bulges into Middle Valley, but i t also merges into Sovanco Ridge in the northwest (Profile e-e'--Fig. 21). Unit A strata overlie the basement in the region of mergence, but the maximum water depth is 2330 m, about 140 m shallower than the adjacent floor of Middle Valley. Hence, at the present time, this saddle prevents turbidity currents from entering West Valley from the northeast. West Valley West Valley is a deep trough cut off in the north by the saddle between West Ridge and Sovanco Ridge. Its eastern wall is the partly fault-controlled flank of West Ridge, whereas its western wall consists of a series of step-faults (Profiles H-H', D-D1, C-C"--Plates VII, IV, III; Profile f-f'--Fig. 21). These fault blocks are capped by Unit A, which also lies in West Valley. The topographically high region west of the valley is mostly covered by perched Unit A strata, except for the protruding basement ridge crossed by Profile I-I' (Plate VIII). Hence, this region appears to have undergone general uplift relative to the abyssal plain to the west, where the top of Unit A is deeper and i t is overlain by Unit B. Part of the sediment on the southern part of the region is Unit C, derived from Heck Seamount chain to the south. It is suggested that West Valley and the western flank of West Ridge may be the locus of most recent rifting and extrusion near the northern end 67 of Juan de Fuca Ridge, developed on an older area of sea floor, with accompanying block faulting and uplift. West Valley terminates abruptly against the eastern end of Heck Seamount chain and the northwestern flank of Endeavour Seamount. This is in line with the southern termination of West Ridge, and may be a dextral transform fault zone between possible spreading loci West Valley-West Ridge and Endeavour Trough. More profiling is necessary to confirm the existence of this fault. Ridge Crest Near 48°N Latitude Endeavour Seamount and adjacent Endeavour Trough are devoid of sediment (Profile B-B1—Plate II). At least in places the eastern flank of Endeavour Trough is probably a fault with 190 m of vertical displacement (Profile A-A1--Plate I). The lack of sediment and the appearance of Endeavour Trough compared to adjacent areas suggest that i t may be the locus of recent rifting and extrusion. A low basement ridge forms the western margin of the crest of Juan de Fuca Ridge (Profile A-A'). The basement in the southern part of the study area is generally high, lacking in relief compared to Juan de Fuca Ridge to the north, and nearly devoid of sediment except on the eastern flank where sediment onlaps from Cascadia Basin (Profile A-A'). This region is apparently typical of much of Juan de Fuca Ridge south of the study area. Summary Juan de Fuca Ridge north of about 48°20'N is different from the ridge south of that latitude in its greater relief. Near the end of Unit A time (sometime less than 0.7 million years ago) turbidity currents were able to 68 spread over the ridge crest, except for a small area on Middle Ridge and most of West Ridge (Fig. 16). Nevertheless, East, Middle, and West Ridges and the intervening valleys were present as topographic features, although of generally subdued relief. Approximately at the end of Unit A time, a period of deformation occurred, involving block faulting and uplift and tilting of East, Middle, and West Ridges. Subsequent Unit B deposition was controlled by this secondary relief and hence confined to East and Middle Valleys. 2. Sovanco Fracture Zone The Sovanco Fracture Zone follows an irregular but elevated basement feature herein called Sovanco Ridge (Fig. 15, 17) which is to a large extent controlled by faulting (Fig, 16). South of Sovanco Ridge on Juan de Fuca Abyssal Plain the depth to basement is about 3350 m, 500 m deeper than i t is under most of Cascadia Basin north of the ridge (McManus et a l , 1972). Hence, Sovanco Fracture Zone is a major tectonic feature. The crestal region of Sovanco Ridge in the study area consists of two high areas on which basement is exposed, separated by a "valley" in which basement is deeper and sediment-covered (Fig. 18). Although faulting apparently has played a major role in the development of Sovanco Ridge, the occurrence of features which look like volcanic peaks (Profile L-L'--Fig. 23) from which pillow fragments and glassy basalts were dredged suggests that extrusive volcanic activity has also occurred. The southern flank of Sovanco Ridge is cut by a series of faults which are most obvious south of the eastern crestal high area (Profiles H-H', I-I'--Plates VII, VIII) where they intersect the north- and northeast-trending fault system of Juan de Fuca Ridge (Fig. 16). Faults cut basement and the overlying Unit A. Profiles to the west also show faulting, but more prominent FIGURE 23. a. Line-drawing interpretation of seismic reflection profile L-L'. Location of profile and key-to symbols shown in Figures 51 and 53 in Appendix A. b. Line-drawing interpretation of seismic reflection profile M-M', with coincident magnetic anomaly profile. Location of profile and key to symbols shown in Figures 51 and 53 in Appendix A. 70 is the tilting of Unit A (Profiles J-rJ\ f J T K ' , K-K1—Plates IX, X, XI; Profiles L-L', M-M'--Fig. 23). Only line-drawing interpretations of L-L' and M-M* are shown because the quality of the original records is too poor for reproduction. Profile K-K' crosses the southern end of what may be a channel allowing sediment supplied by Revere Channel to the north to cross Sovanco Ridge and reach the abyssal plains to the south. Tilting, faulting, and uplift of Unit A indicates that uplift of the Sovanco Ridge region has occurred, but prior to deposition of Unit B. Although vertical displacement has apparently occurred on the faults of the Sovanco Fracture Zone, i t cannot be determined from the seismic profiles whether or not horizontal displacement has also occurred. Some of the apparent vertical displacement could be the effect of juxtaposing sea floor of varying levels along strike-slip faults. Along the northern flank of Sovanco Ridge, the basement has a dip to the north and is overlain by Unit B; some faulting has occurred, but i t is generally less prominent than on the southern flank (Profiles J - J ' , J-K'. L-L', M-M'). The contact between Units A and B cannot be definitely identified, although i t can be identified farther north in Cascadia Basin (McManus e_t a l , 1972). A buried basement "moat", perhaps fault-controlled, may follow the northern edge of Sovanco Ridge (Fig. 17). The eastern margin of Sovanco Ridge is the steeply dipping fault described previously as the western wall of Middle Valley, on which the base-ment relief is at least 2000 m (Profile G-H--Plate VII; Profile d-c'--Fig. 20). 3. Seamount Chains Few seismic profiles were run over the seamount areas during this study, However, because l i t t l e or no sediment cover occurs over the igneous 71 basement on the seamounts (for example, Profile N-N'--Plate XII) bathymetric contours are essentially basement contours along the two seamount chains (Fig. 15, 17). The generally flat-lying sediments of the abyssal plains onlap the bases of the seamounts from the north, south, and west (Profiles L-L', M-M'— Fig. 23). However, north of Heck Seamount, the basement is covered by an irregular layer of sediment (Unit C); this is about 275 m above the level of the abyssal plain sediment and is probably derived largely from Heck Seamount (Profile I-I'--Plate VIII). The basement over the seamount exhibits hyperbole and "tails", characteristic of rough, irregular, and steep igneous topography. Profile N-N' (Plate XII) crosses from the eastern peak of the Heck Chain over the eastern end of the Heckle Chain to a seamount just south of the study area. Between Heck and Heckle Chains the profile crosses a region of high, irregular basement, in places overlain by a veneer of sediment (Unit C) which is about 130 m above the sediment level on the abyssal plain to the west and probably largely locally derived from adjacent seamounts. South of the Heckle Chain the profile crosses the eastern end of the abyssal plain, but sediment here is probably also partly derived from adjacent seamounts. Profile A-A1 (Plate I) also crosses the eastern end of the Heckle Chain, as well as the low ridge marking the western margin of Juan de Fuca Ridge. Minor sediment in the basement lows of this profile is probably also Unit C. 4. Abyssal Plains The ability of deposition from turbidity currents to smooth an irregular sea floor to a similar level over a wide area can be demonstrated on Juan de Fuca Abyssal Plain. The depths to the sea floor, to igneous 72 basement, and to Unit A (if identifiable) for 18 locations (plotted on Fig. 24) in the study area are listed in Table I. These data were obtained from Profiles I-I', J-J', J-K'--Plates VIII, IX, X; Profiles L-L', M-M'--Figure 23; Profiles e-e', f-f--Figure 21; and other profiles from the study area not reproduced in this thesis. The depths to the sea floor are remarkably uniform--about 2780 m (uncorrected)--on the abyssal plains. The regions of higher level already mentioned north of Heck Seamount and between the eastern ends of the Heck and Heckle Seamount Chains are at varying water depths (Locations 1, 2, 14, 15). The depths to igneous basement under the abyssal plains are variable, but are deeper in the region north of the Heck Chain than in the region between Heck and Heckle Chains. Locations 8 and 12 are about the same distance from Juan de Fuca Ridge, but, although the water depths are the same, the depth to basement is about 270 m greater at Location 8 between Heck Chain and Sovanco Ridge than at Location 12. This difference may be related to the postulated former presence of the Sovanco Fracture Zone, trending east-west in the region north of the Heck Chain (see discussion on p. 188). Because of the higher level basement, the sediment cover is thinner in the region south of Heck Chain, and also south of Heckle Chain, than in the region north of Heck Chain. Since the depth to the sea floor is so constant, regardless of basement configuration (unless the basement is less than about 2780 m deep) the bulk of the sediment could not be pelagic or locally derived, but must have been derived from the continent and transported into the area by turbidity current processes, and is considered to be Unit B. Unit A is possibly identifiable at Location 12 where the upper surface is about 2925 m deep. Where the basement is at 2880 m farther to the southeast (Location 13), i t does not seem to be present. Since i t is FIGURE 24. Locations of points used in Table I to illustrate depths to sea floor, igneous basement, and the top of Unit A on Juan de Fuca Abyssal Plain and adjacent features. Base map is from Figure 15. 74 Table I. Depths to sea floor, igneous basement, and the top of Unit A at abyssal plain and associated locations. The locations are indicated on Figure 24. Depths are based on a velocity of 1500 m/sec in water and sediment. Location Sea Floor Igneous Basement Top of Unit A 1 2560 m 2630 m -2 2740 2925 2740 m 3 2790 2945 2835 4 2770 3063 2890 5 2790 3020 2890 6 2770 3290 3020 7 2780 3290 3070 8 2780 3290 3000? 9 2780 3290 2925? 10 2760 3245 2955 11 2770 3109? 2965 12 2780 3020 2925? 13 2760 2880 -14 2595 2650 -15 2650 2650 -16 2780 2835 -17 2780 2835 -18 2862 3200 2862 75 unlikely that tectonic processes have affected the basement in that region after deposition of Unit A, i t is probable that the depth to the top of the unit was originally about 2925 m. This coincides with the depth to the top of Unit A at Location 9 north of the Heck Chain. Hence, 2925 m is considered to be an approximation of the depth to the sea floor on the abyssal plain areas at the end of Unit A time. On this basis, considerable tectonism has occurred since that time in the region between Heck Seamount Chain and Sovanco Ridge because the top of Unit A is at varying levels. This probably applies as well to the southern flank of Sovanco, and the high region between the abyssal plain and West Valley because their Unit A sequences were probably continuous with those of the abyssal plain. The top of the sequence in West Valley is at 2862 m; considering the distances involved, the similarity of this depth to the depth to Unit A on the abyssal plains supports the interpretation that West Valley and the abyssal plains were connected during Unit A time. In summary, i t appears that Unit A was deposited in the northern part of Juan de Fuca Abyssal Plain only where the basement was deeper than about 2925 m, the probable depth to the top of Unit A at the end of Unit A time. The uniformity of the depth to the top of this unit everywhere on the abyssal plain where i t was deposited is likely by analogy with the present uniform depth to the top of Unit B. On this basis, considerable tectonism occurred at the end of Unit A time and early in Unit B time to result in the varying depths to Unit A presently observed in the region between Heck Seamount Chain and Sovanco Ridge. This does not imply that the sea floor on Juan de Fuca Abyssal Plain and the sea floor on Juan de Fuca Ridge were at the same level at the end of 76 Unit A time. The latter was probably at least 225 m shallower (see p. 192). A similar situation is presently observed north and south of Sovanco Ridge (Fig. 23b). However, West Valley was apparently at the abyssal plain level rather than the shallower level of Juan de Fuca Ridge and Cascadia Basin. 5. Cascadia Basin During this study, no seismic profiles which show penetration to igneous basement across Cascadia Basin were obtained. However, three published profiles cross the basin from the continental slope to Juan de Fuca Ridge (Ewing et a l , 1968; Hayes and Ewing, 1970; McManus et a l , 1972) and show the basement dipping to the east at an average angle of 0.5° from East Ridge to the base of the continental slope. The depth to basement in Cascadia Basin at the base of the slope in the southern part of the study area, as inferred from these profiles, is at least 3800 m, and the thickness of the overlying sediment is about 1500 m. However, near 49°N latitude in the northern part of the study area the depth to basement and sediment thickness are less--3675 m and 1125 m respectively (Profile 0-0'--Plate XIII). These are a l l minimum values, using a sound velocity of 1500 m/second in both water and sediment. If estimates for increasing sound velocity with increasing sediment thickness (from Moore, 1968) are considered, then the true sediment thicknesses are probably about one third greater, and the depths to basement about one eighth greater than the values presented. None of the seismic profiles show penetration to igneous basement below the continental slope itself. The basement under Cascadia Basin is smooth and featureless compared to the basement on the crest and western flank of Juan de Fuca Ridge. There apparently are no seamounts. The largest irregularities observed on the 77 profiles (McManus et a l , 1972, Fig. 11) appear to be tilted fault blocks similar to those on East and Middle Ridges, but with a maximum relief of 450 m. This is in agreement with the gravity data (Dehlinger ejt a l , 1970) which indicate that there is no highly irregular basement topography in Cascadia Basin (Fig. 7). Sediments, generally acoustically transparent, which immediately overlie the igneous basement in these profiles, have been interpreted to be pelagic sediments (Hayes and Ewing, 1970). However, the better-quality seismic profile of McManus ejt al_ (1972), Figure 11, shows that the sequence contains many weak reflectors and that the sediments do not drape the basement features in the characteristic manner of pelagic deposits. These data, together with the difficulties encountered in the pelagic interpretation (discussed in detail by Hayes and Ewing, 1970), led McManus et al to conclude that the sequence consists predominantly of turbidites. Like the basement, deeper reflectors show a dip to the east, which decreases upward in the section. The uppermost reflectors of the sedimentary deposits are horizontal, or even slightly westward-dipping on Nitinat Fan. The lower part of the sequence is deformed with the basement, and can be traced into the tilted Unit A section on East Ridge (McManus et a l , 1972, Fig. 11). The eastward dip of the sedimentary section and the apparent increase in thickness of every bed towards the continental slope, as noted by Hayes and Ewing (1970), could be partly due to compaction (estimated to be in the order of 40% in a section of this thickness, after Moore, 1968). However these effects could also be caused by progressive tilting of the basement and essentially continuous turbidite deposition. 78 The portion of Cascadia Basin north of the study area has been described by McManus e_t a_l (1972) . The igneous basement surface is smooth and gently undulating. It is overlaid by a sedimentary section of varying thickness, the lower part of which is partly acoustically transparent and semi-parallel to the basement surface, whereas the upper part is generally flat-lying (the Lower and Upper Turbidites of McManus et a l , considered to be equivalent to Units A and B of this thesis). 6. Continental Slope The seismic profiles reveal that the continental slope consists of a series of discontinuous ridges and blocks, often apparently fault-bounded, behind which sediments have been ponded with the effect of subduing the relief. These ridges and blocks are generally composed of acoustically reflective material, apparently sedimentary, acting as acoustic basement as previously described. Overlying the ridges and blocks, or ponded behind them, are well-stratified sediments containing strong reflectors. In Profile 0-0' (Plate XIII) from the northern part of the study area, the igneous basement can be traced from Cascadia Basin under the con-tinental slope. The basement is dipping to the east at an angle of about 1 . 7°, which may increase just east of the base of the slope. Alternatively, this may be only a local step in the basement, as the basement can be traced only a very short distance east of this position. The igneous basement does not appear to be involved in the deformation affecting the overlying sedimen-tary section. The nearly flat-lying reflectors of the Cascadia Basin sedimentary section increase in dip with depth, the deepest reflectors being essentially parallel to the igneous basement, and the uppermost reflectors horizontal. 79 Abruptly, probably at a fault, the flat-lying and gently dipping section is truncated. East of this, what is apparently the same section is folded and faulted in an uplifted, anticlinal ridge. The penetration and resolution of reflectors in this structure are much poorer than in the undeformed section to the west. The signal obtained from the igneous basement is also much weaker. About 2 km to the east, what is probably again the same sedimentary sequence is more intensely folded and faulted, and penetration into the structure is very limited. It acts as an acoustic basement, and igneous basement cannot be traced under this structure. The structure has acted as a dam for sediment for a considerable time, ponding behind i t a section about 750 m in thickness. The deepest part of this ponded section is slightly deformed, suggesting that i t was accumulating during the formation of the dam structure. The ponded section overlies acoustic basement. It appears that a process of continental accretion occurs at the base of the continental slope, in that Cascadia Basin turbidite deposits are folded and up-lifted to form ridges which dam sediments moving down the continental slope, and hence the continental slope is built outwards. This process can be clearly seen in progress at the base of the slope as in Profile 0-0'. However, older regions of the slope have been further modified, by submarine erosion and continued deformation, fault movement, and uplift, and the picture is not as clear. The level surface behind the dam in Profile 0-01 is the western part of a broad plateau formed by damming of sediments by acoustic basement highs, the one just described and others to the south. These features are obvious on the bathymetric chart (Fig. 15, p. 287) between 48°45' and 49°00'. Profile Q-Q' (Plate XIV) also crossed this plateau. Although the profile lacks the resolution and penetration of Profile 0-0', the basement and overlying 80 structures can be discerned. The upper continental slope east of the plateau consists of well-stratified, prograding sediments draping blocks of acoustic basement which appear to be fault-controlled. Profile P-P' (Plate XIII) crosses from Cascadia Basin over the base of the continental slope. The deformation in the structure at the base of the slope is more intense than in Profile 0-0*, but the Cascadia Basin deposits again appear to be truncated at a fault. Faulting is also inter-preted to have played a major role in forming the rugged topography of the remainder of Profile P-P'. South of the region of Profile P-P1, a broad canyon, which probably carries sediment supplied from the prograding upper continental slope north of Clayoquot Canyon (Fig. 15), crosses the lower continental slope. Profile R-R' (Plate XV) crosses this broad unnamed canyon. The southeastern wall is steep and may be fault-controlled. The northern wall is the high feature forming the rim of the northern plateau as previously described. Profile R-R also crosses this plateau and the lower slope structures but the quality of that part of the profile is very poor. On the upper continental slope, Profile R-R1 crosses shallow Clayoquot Canyon (which does not appear to be connected to the previously described canyon) and several narrow canyons, including Father Charles. Underlying the upper slope is acoustic basement with considerable relief in-filled by layered sediments which show effects of deformation. The upper slope canyons truncate the stratified sedimentary section, and may be partly controlled by faulting. Profile S-S' (Plate XVI) crosses from Cascadia Basin over the con-tinental slope to the continental shelf in the central part of the study area The flat-lying Cascadia Basin sediments are terminated at a high ridge compos of acoustically reflective material whose western flank is apparently fault-81 controlled. The trough east of the ridge contains well-stratified, slightly-deformed sedimentary material. The structures underlying the middle slope in this profile are difficult to interpret. However they appear to be acoustic basement blocks infilled and draped by sediment, as described under the slope to the north. It is probable that faulting and deformation of the basement blocks continued during deposition of the stratified sediment. The stratified sediment merges with similar deposits underlying the continental shelf, which are considered to be Pliocene in age (Tiffin e_t al_, 1972). Unconforrnably overlying these strata at the shelf edge is a wedge of sediment thought to be Pleistocene in age (B. E. B. Cameron, pers. comm. 1972). Profile T-T1 (Plate XVII) is a more detailed crossing of the ridge at the base of the slope crossed by Profile S-S'. The igneous (oceanic) basement is visible at a depth of about 2000 m at the base of the slope. It is dipping to the east at an angle of about 1 . 2°. Overlying this basement is a section of Cascadia Basin strata containing strong reflectors. The reflectors lose their coherence abruptly at the western flank of the ridge which is the basal structure of the continental slope. Some reflectors--apparently faulted, folded, and tilted--may be traced for short distances into the ridge structure. The igneous basement can be traced only about 2 km under the structure, and does not appear to be involved in the structure. The eastern flank of the ridge is probably also fault-bounded, but side echoes make interpretation difficult and ambiguous. East of the trough behind the ridge, the slope rises steeply to a plateau, under which strata drape and i n f i l l an irregular acoustic basement surface, as observed in the mid-slope area on other seismic profiles. As is evident from the bathymetry (Fig. 15) the structures at the base of the continental slope do not form continuous topographic features. 82 However, seismic profiles (not reproduced in this thesis) south of Profile T-T' show versions of varying reliefs of the same structures--apparently formed by anticlinal folding, faulting, and uplift of Cascadia Basin sediments--which have been described on Profile T-T' and those to the north. This process until at least very recently, i f not at present, has occurred at the base of the continental slope everywhere in the study area. Profile U-U' (Plate XVIII) crosses the lower continental slope about 55 km southeast of Profile T-T1. The nearly flat-lying turbidites of Cascadia Basin are folded, faulted, and uplifted, possibly along high-angle reverse faults. Most fault planes are depicted to be vertical, since the true configuration cannot generally be determined from the profile. The trough behind the westernmost ridge contains nearly flat-lying sediments, probably deposited after most of the folding and uplift had occurred. The sediments at the eastern edge of Cascadia Basin are arched as i f in the preliminary stage of formation of a structure like the anticlinal ridge to the east. The high ridge composed of acoustic basement in Profile U-U' is probably fault-bounded, and is considered to represent an advanced stage of the process. Dredging on this feature recovered mudstone and siltstone, evidence for the sedimentary composition of the acoustic basement on the continental slope (see p. 173). The middle and upper continental slope in this southern part of the study area are cut by large submarine canyons. Profile V-V (Plate XIX) follows the southern flank of Nitinat Canyon, crosses Nitinat Canyon, then follows the southern flank of Barkley Canyon. As in other profiles across the middle continental slope to the north, the acoustic basement, showing consider-able relief and major faulting, is in-filled and draped by a gently deformed sedimentary section containing strong reflectors. 83 In summary, the structures under the continental slope appear to be the result of a continuing process caused by relative compression between Cascadia Basin and the continental slope involving sequential uplift of fault blocks or thrust slices. Additional evidence for this process was provided by a hole drilled by JOIDES into an apparently similar structure on the lower continental slope west of Oregon. It encountered a turbidite section below 120 m similar to the abyssal plain turbidites drilled below the Astoria Fan deposits (Von Huene e_t a l , 1971). Palaeobathymetric indicators in the lower slope section suggest that i t was deposited in water 500 m deeper than its present depth. In the study area, processes of erosion and deposition modify the primary structures, particularly on the middle and upper slope. Summary of Structural Interpretations The distribution of overlying sediment indicates that the elevation and relief of the northern end of Juan de Fuca Ridge and at least the eastern half of Sovanco Fracture Zone are greater now than in the past. The most recent turbidite sedimentation (Unit B) has been confined to lowlands at either the Cascadia Basin or Juan de Fuca Abyssal Plain levels. In contrast, near the end of Unit A time, turbidites were deposited over much of Sovanco Ridge, Juan de Fuca Ridge, and the region west of West Valley, which are now elevated above the level of turbidite sedimentation. During Unit A time there were probably also two levels of turbidite deposition, the Cascadia Basin and Abyssal Plain levels. West Valley and the southern flank of Sovanco Ridge were at the latter level. The tectonics of the vicinity of the northern end of Juan de Fuca Ridge are related to uplift, as well as tension on the ridge crest it s e l f . 84 The coincidence of uplift and deformation with the apparent change in sediment type at Unit A-Unit B contact may be fortuitous, as i t is difficult to corre-late except by chance local deformation on Juan de Fuca Ridge with a change in continental sediment supply. Alternatively, the change in appearance of reflectors from Unit A to Unit B may be partly due to the effects of deforma-tion on Unit A, rather than solely a change in sediment type. On the continental slope, compression is resulting in folding, faulting, and sequential uplift of Cascadia Basin deposits. The net result is outbuilding of the slope and thickening of the section above the oceanic basement. MAGNETIC DATA Introduction The total magnetic field was measured using a Varian proton precession magnetometer during four of the cruises to the study area, and consequently magnetic data were obtained with most of the seismic profiles. The magnetic fields measured at each of eight intersections of profiles representing a l l four cruises were compared, and the absolute mean deviation was 14 gammas. The main causes of the differences in readings at these intersections are thought to be navigational errors and neglect of diurnal variations. Because the total intensity of the regional geomagnetic field increases by about 1000 gammas from southwest to northeast across the study area, approximate values for the regional field* were subtracted from the total magnetic field data, to give residual magnetic anomalies. regional field was calculated using Chart 1965.0. F-Isodynamic Chart, Canada--Lines of Equal Total Magnetic Intensity and Equal Annular Change, published by the Department of Mines and Technical Surveys, Dominion Observatories Branch. 85 Primarily because of greater density of profiles and more accurate navigation, the magnetic data of Raff and Mason (1961) for much of the study area is superior to that collected during this study. Except in eastern Cascadia Basin and on the continental slope, the new data were not contoured, but are plotted above the coincident seismic profiles (Plates I to XIX and Fig. 23b). Comparisons between the data collected during this study and that of Raff and Mason (1961) reveal striking similarities (Fig. 25). The differences are due to navigational errors and the different techniques used in removing the regional field. However, the overall anomaly pattern obtained by Raff and Mason was duplicated during this study. The main features of the magnetic anomalies in the study area are noted in the following section. Juan de Fuca Ridge a. A lack of correspondence exists between the basement topography and the linear magnetic anomalies in many places on the crest and flanks of Juan de Fuca Ridge (for example, Profile D-D'--Plate IV). This has been noticed on other portions of the ocean ridge system. Raff and Mason (1961) noted that a statistical correlation analysis made of the magnetic and topographic relief for some of the area between 40° and 52°N latitude off the west coast of North America showed that topographic ridges and magnetic "highs" correspond in only 65% of the cases. Menard and Mammerickx (1967) noted that on several parts of the East Pacific Rise, including Juan de Fuca Ridge, topographic trends are parallel to trends of magnetic anomalies but are not specifically related. They considered this to indicate that both hills and anomalies originate at the crest of the rise, and noted that the hills are associated with relatively rapid spreading (based on the magnetic 86 FIGURE 25. Comparisons between residual magnetic anomaly profiles obtained with seismic reflection profiles D-D' and E-E', and the Raff and Mason magnetic anomalies along the same lines. 87 reversal time scale) in contrast with the mountainous topography and deep rifts which are associated with slower spreading. For the Mid-Indian Ocean Ridge Udintsev et al (1971) calculated that the magnetic effect of irregular topography should generate anomalies of amplitude 400 to 700 gammas, which are of the same order as the ones observed, and yet correlation of anomalies and topography is not clear. No satisfactory explanation for these observations is available. It has been suggested that the lack of correspondence between bathymetry and anomalies could be explained i f the depth to anomaly source were greater than generally assumed (Van Andel and Bowin, 1968). This seems unreasonable because of the high magnetic intensity of basalt dredged from the ocean floor. Watkins and Richardson (1968) demonstrated that the magnetic profile of the Mid-Atlantic Ridge at 22.5°N can be duplicated by a model which takes into account the topography. However, i t has not been demonstrated that such a model can be constructed for other regions such as that discussed by Udintsev e_t al_ (1971). Nevertheless, the conclusion of Watkins and Richardson that the commonly used model employing flat prisms of alternating polarities is inconsistent with the observed structures of ocean ridges seems valid. It should be kept in mind that the origins of the magnetic anomalies may be much more complex than most models indicate, particularly in areas of con-siderable basement relief. b. Large anomalies occur in western Cascadia Basin where there are no basement irregularities. This further demonstrates that the linear magnetic anomalies are in many places independent of topography. c. A broad positive anomaly is associated with the crest of Juan de Fuca Ridge in the southern part of the study area (Profile A-A'--Plate I). This is clearly the broad central anomaly (representing the Brunhes Magnetic 88 Epoch) which follows the crest of Juan de Fuca Ridge (Fig. 3), and indicates the spreading centre. This broad positive anomaly branches at approximately the latitude of the northern end of Endeavour Trough, the eastern branch following Middle Ridge and the western branch West Ridge and West Valley (Fig. 4). The negative anomaly between these branches is associated with Middle Valley in the north but extends south of the topographic expression of Middle Valley to the latitude of Endeavour Trough. A similar splitting of the central anomaly is observed near the northeastern end of Explorer Ridge where i t is thought to be the result of two spreading centres which have perhaps functioned alternately (A. G. Thomlinson, pers. comm., 1972). If Middle Ridge has acted as a spreading centre, i t apparently was early in the Brunhes Epoch because the ridge is blanketed by Unit A sediments. The magnetic data combined with bathymetric, seismic reflection, and petrologic data indicate that the western flank of West Ridge and West Valley are the loci of most recent sea floor spreading in the northern part of the study area. In the southern part of the area, Endeavour Trough is the locus of activity, and a short dextral transform fault may join these features. A discussion of the evidence leading to these conclusions is presented in Chapter V. d. South of Endeavour Trough, the linear magnetic anomalies are symmetrical in a direction perpendicular to the ridge crest. In contrast the anomalies near the northern end of the ridge are not symmetrical in that those of ages 0.7 to about 5 million years are absent on the western flank. This lack of symmetry is attributed to recent activity along old east-west trends north of Heck Seamount Chain, and is further discussed in Chapter V. 89 Sovanco Fracture Zone and Cascadia Basin to the North The anomaly pattern associated with Sovanco Fracture Zone is irregular but generally trends parallel to the topographic expression of the fracture zone (Fig. 4). Much of the main positive anomaly lies north of the topographic relief of Sovanco Ridge. Anomalies in the region farther to the north are irregular and generally non-linear. At its. eastern end the main positive anomaly of Sovanco may merge with the positive anomaly of West Ridge. Profiles across Sovanco show many small, local magnetic anomalies which may be caused by localized igneous activity (Profiles M-M'—Fig. 23; Profiles J-K1, 1-1'— Plates X, VIII). Abyssal Plains A strong positive anomaly trends nearly east-west between Sovanco Fracture Zone and the Heck Seamount Chain (Fig. 4). Profiles across this region (Profiles M-M', I-I') show no topographic feature associated with the anomaly. It is postulated to mark a former position of Sovanco Fracture Zone (see p. 188). South of the Heck Chain, the northeast-trending regional anomalies cross the abyssal plains. Seamount Chains The linear magnetic anomalies generally cross the seamount chains with only minor distortion (Fig. 4). The peaks along the seamount chains have magnetic "highs" associated with them of the same sign as the anomaly in which they occur (Raff and Mason, 1961). This implies that most seamount construction occurs during formation of the linear anomalies at the crest of Juan de Fuca Ridge, and the seamounts are then carried away by sea floor spreading. A few embayments of reversed polarity (for example, Figure 4, 90 about mid-way along the Heckle Chain and on Heck Seamount) suggest that eruptions in some places continue for some time after the seamount is carried off the ridge crest, and in sufficient quantity to change the sign of the feature. Continental Slope and Eastern Cascadia Basin Magnetic anomalies under Cascadia Basin are high-amplitude, short wave-length anomalies. Under the continental slope the anomalies change to low-amplitude and long waVe-length. A chart of contoured magnetic anomalies prepared from data collected during this study illustrates the difference (Fig. 26). The ages of the anomalies on Figure 26 are extrapolated from Vine's (1968) interpretation of Raff and Mason's data (Fig. 3). Because of the different techniques used to remove the regional f i e l d , the contour values do not correspond but the anomalies can be easily correlated. The change to low-amplitude, long wave-length anomalies may be at least partly related to the dipping of the magnetic anomaly source layer beneath the continental slope, and hence an increasing source depth. Anomaly source depth determinations were carried out on the anomalies using three different graphical methods. The values obtained using the "straight-slope" method, which is based on the empirical relationship between the depth to the causative body and the horizontal extent of that part of the anomaly that has a uniform slope (Nettleton, 1962; Hood, 1965), are plotted on Figure 27. The depths obtained using the other two methods—the "half-slope" method (Peters, 1949) and the "half-width" method (Dobrin, 1960)--were consistently deeper than those from the "straight-slope" method. It is expected that many of the calculated depths have large errors. However, at locality A (Fig. 27) the depth to igneous basement (the probable source of the magnetic anomalies) FIGURE 26. Magnetic anomalies under eastern Cascadia Basin and the continental slope between 48° and 49°N latitude contoured from magnetic data recorded during the present investigation. The ages of the anomalies are inferred from comparison with Vine's (1968) interpretation of the Raff and Mason (1961) anomalies. FIGURE 27. Magnetic anomaly source depths under eastern Cascadia Basin and the continental slope between 48° and 49°N latitude, calculated by the graphical "straight-slope" method (Nettleton, 1962; Hood, 1965). 93 estimated from the seismic profile 0-0' is 3.7 km, compared to the calculated source depth of 3.6 km. Although somewhat irregular, the trend of increasing anomaly source depth under the continental slope is obvious from Figure 27, and indicates a dip of more than 10° for the source layer. Another factor becomes important as the sediment thickness and depth to anomaly source increase: the temperature of the source layer. The blocking temperatures (the temperatures at which remanent magnetization is lost) determined for basalts from the Mid-Atlantic Ridge near 45°N are as low as 100°C and values of less than 300°C are common, although temperatures tend to increase away from the median valley (Irving et al , 1970). A Curie temperature of 175° has been measured in a basalt pillow from Juan de Fuca Ridge at 46°N (Marshall and Cox, 1970b) and the blocking temperature could be lower by 50° or more (Irving et a l , 1970). A mean geothermal gradient of 1.10 x 10"3oC/ cm has been measured near the base of the continental slope west of Oregon (Korgen et a l , 1971). Using this result, then the temperature at the top of the igneous basement at the eastern edge of Cascadia Basin in the study area is already approaching 200°C. Hence the changing character of the anomalies may be partly related to decrease in susceptibility representing the loss of thermo-remanent magnetization. Silver (1971a) reproduced the anomalies observed under the continental slope west of northern California with the aid of a computer. The best-fit model required a dipping basement surface and a marked susceptibility decrease beneath the slope. Silver suggested that the decrease in susceptibility could represent either loss of thermo-remanent magnetization or mechanical disruption of the basement beneath the slope. 94 Discussion and Summary of Magnetic Data The linear magnetic anomaly pattern in the northeast Pacific Ocean is the major evidence that Juan de Fuca Ridge is an active spreading centre. When examined in detail in the study area, the magnetic anomalies show con-siderable deviation from linearity parallel to the ridge. That the process of sea floor spreading be complicated near a transform fault zone is perhaps reasonable, particularly i f , as in the case of Juan de Fuca, the spreading direction has recently changed. The presence of the broad positive anomaly trending east-west north of Heck Seamount Chain suggests that i t might mark a former location of the Sovanco Fracture Zone. Seismic reflection profiles across the area of the anomaly show no evidence for a fossil fracture zone (Fig. 23), but the igneous basement is deeper than i t is under Juan de Fuca Abyssal Plain south of the Heck Chain (Table I). It is apparent from a comparison of magnetic anomalies and basement topography that the anomalies generally developed independently of igneous basement structure. Nevertheless, the general parallelism of these linear characteristics indicates that both originate at the spreading centre. The reason why major secondary changes in topographic relief apparently do not greatly affect the primary magnetic anomalies is not understood. The magnetic reversal time scale indicates that the oldest igneous basement observed on the seismic reflection profiles in the study area is that near the base of the continental slope which is about 6 million years old. Possible evidence of subduction is provided by the observation that the linear magnetic anomalies of the ocean floor can be traced under the continental slope. The basement source of the anomalies dips below the slope at an angle of more than 1 0°. The decrease in amplitude and increase in wave-length of the anomalies under the continental slope is probably related both 95 to the increasing depth to the basement source layer and the loss of thermo-remanent magnetization as the temperature increases, although mechanical disruption of the source layer is also a possible cause. If i t could be demonstrated that the ages of the magnetic anomalies under the continental slope are in places younger than the ages of overlying sedimentary deposits then this would be definite evidence for subduction. Unfortunately, the prograding structure of much of the continental slope in the study area prevents sampling by dredging of anything but young sediments, and the ages of units observed on the seismic reflection profiles are not known. Drilling on the continental slope in an area where the ages of underlying anomalies are clear would undoubtedly provide valuable data. CHAPTER IV SAMPLES FROM THE OCEAN FLOOR INTRODUCTION Forty-one dredge hauls were attempted in the study area, and samples of the bedrock were obtained in seventeen. Basalt was collected in thirteen hauls from the ridges comprising the crestal area of J u a n de Fuca Ridge, from Sovanco Ridge, and from Heck and Heckle Seamount Chains (Stations 1 to 13, Fig. 28), whereas mudstone was dredged in four hauls from the continental slope (Stations 14 to 17, Fig. 29). In addition, unconsolidated sedimentary material was sampled in nine hauls from the ridge area and six hauls from the continental slope. Summary descriptions of the seventeen dredge hauls in which bedrock was obtained are given in Appendix B (Table XIII). Locations and brief accounts of the other twenty-four hauls are also presented in Appendix B (Table XIV). DREDGING TECHNIQUE Dredging was done from the well-decks of both the C.S.S. PAHIZEAU and the C.F.A.V. ENDEAVOUR 3 on the PARIZEAU through an A-Frame on the star-board side, and on the ENDEAVOUR through a complicated arrangement over the bow, except on Cruise 71-23, when an A-frame was available on the well-deck of the ship. Two types of dredges were used. The pipe dredge (Frontispiece, middle right) usually retained unconsolidated material as well as bedrock, FIGURE 28. Locations of dredge stations in the Juan de Fuca Ridge area in which samples of bedrock (basalt) were recovered. Arrows indicate direction and approximate path of dredge. Base map is from Figure 15. 98 FIGURE 29. Locations of dredge stations on the continental slope in which samples of bedrock (mudstone) were obtained. Arrows indicate direction and approximate path of dredge. Base map is from Figure 15. 99 and in many hauls probably f i l l e d up with unconsolidated material upon first contact with the bottom, leaving no room for rocks subsequently encountered. Bottom photographs (C. B. Lister, pers. comm., 1972) and seismic reflection profiles indicate that, in the study area, many outcrops are surrounded by sediment-covered lowlands, and it is difficult to control positioning of the dredge accurately enough to avoid the sediment-mantled regions. Also, the resolution of the seismic profiles is not fine enough to indicate whether bedrock is exposed or mantled by thin or patchy sediment. The chain bag dredges (Frontispiece, lower left) did not generally retain unconsolidated material and were more effective than the pipe dredges on areas of basaltic bedrock. However, on the continental slope the small chips of mudstone recovered embedded in the unconsolidated sediment in the pipe dredges would not have been retained by the chain bag dredges. Dredge stations were chosen at sites of bedrock outcrop indicated by seismic reflection profiles, and located by echo-sounding. The ship was usually positioned over the bottom of the slope to be dredged and the dredge was lowered a distance approximately equal to the water depth. The ship then moved in the up-slope direction while additional wire was paid out, until a length equal to approximately one and one half times the water depth was out, or until the tensiometer showed the dredge to be on bottom. The ship continued its course in the direction up-slope, monitored by the echo sounder, until the tensiometer indicated that the dredge was dragging along and biting the bottom, or until the summit of the feature was reached. The ship was then stopped and the dredge hauled in. Usually the best dredging occurred at this time. On several occasions the dredge became lodged on the bottom and i t was necessary to manoeuvre the ship to free i t . Sometimes the ship was pulled ° 100 back in the direction down-slope--then it was possible to infer quite accurately the depth from which the samples were obtained. LABORATORY INVESTIGATIONS All rocks obtained were examined and a large number from each dredge haul were sawn. Fifty-eight representative specimens were selected for study in thin section. Plagioclase compositions were determined on the universal stage (Slemmons, 1962). Powdered samples of fresh and altered basalts, oxide coatings, and oxide nodules and slabs were analyzed by X-ray diffractometer. Thirty-two specimens were chosen for chemical analysis. Samples from both rims and interiors of six of these specimens with distinct altera-tion rims were analyzed. Similarly samples from both the fresh, black glass and the crystalline interior were analyzed from each of four specimens with glass selvages. In addition, a small fragment of glass was analyzed from a dredge haul in which no crystalline basalt was obtained. The samples were fused with lithium metaborate and analyzed by the author by atomic absorption. The method used was devised by the author, based on various techniques described in the literature. The method and the accuracy of the results are discussed in detail in Appendix B, Part 2. The U.S.G.S. Standard Basalt BCR-1 was used as a standard. Elements determined were: Si, Al, T i , Fe, Ca, Mg, Na, K, and Mn. The results are presented as the oxides Si02, AI2O3, T1O2, FeO, CaO, MgO, Na20, K20, and MnO. Age dating by counting of fission tracks was attempted on glass samples from each dredge haul. Several chips of glass from each of twenty-one specimens were cut, ground, and rough-polished to slabs about 1 cm by 1 cm. One chip from each set was irradiated, mounted in epoxy, and finely polished. 101 Two or three others were mounted and polished only. Irradiated and unirradiated polished mounts were etched in 22% hydrofluoric acid at room temperature for 30 seconds. Track counting methods, the reliability of fission track dates in glass, and track identification criteria have been discussed by Fleisher and Price (1964; 1965) and Aumento (1969). The author followed the technique of Aumento (pers. instruction, 1971). No detailed investigation of the mudstone or unconsolidated material obtained in the dredge hauls was done by the author. However, samples were submitted to B. E. B. Cameron (Geological Survey of Canada) for microfaunal studies and the results are presented here (p. 170-174). BASALTS AND RELATED DEPOSITS Petrography of the Basalts Although their mineralogies are similar, the basalts from the study area show great variations in external form, texture, degree of crystal-linity, and extent of alteration. These variations are observed even among basalts from one dredge haul. Detailed petrographic descriptions of the various basalts are given in Appendix B, Part 1, under the separate station headings. Only the principal features are discussed in the following section, and these features are summarized in Table II. Alteration products and vesicle f i l l i n g s , ferro-manganese deposits, and breccias and tuffs associated with the basalts are discussed in separate sections elsewhere in the chapter. Specimens pictured in the photographs are identified by their dredge station numbers followed by their specimen numbers (for example, 4-1). The photographs are arranged so as to group those of specimens from each dredge haul, and the dredge hauls are presented in numerical order. 102 TABLE II. Major petrographic characteristics of principal types of basalt dredged from Stations 1 to 13 JUAN DE FUCA RIDGE 1 Middle Ridge 2 Middle Ridge 3 West Ridge 4 Endeavour Seamount SOVANCO RIDGE 5 East End Glomeroporphyritic plagioclase basalt. A few glass selvages. Groundniass textures generally intersertal. Fig. 30a, b, c. (a) Angular blocks of holocrystalline basalt with intergranular textures. Fig. 30d, e. (b) Fine-grained basalt with glass selvages. Some fragments have breccia coatings. Some are very vesicular. Fig. 30d, f. Fresh porphyritic (in places glomeroporphyritic) olivine basalt. Glassy pillow fragments. Fig. 31a,. b. Fresh glomeroporphyritic plagioclase basalt. Two pieces of a large, S-shaped, glassy pillow. Variolitic texture. Fig. 31c, d, e. 6 Southern flank (a) Angular blocks of hologrystalline basalt with intergranular texture. Fig. 32a, b, c. (b) Glomeroporphyritic plagioclase basalt, generally with glass selvages. Fig. 32d, e, f. (c) Fine-grained basalt. Glassy pillow fragments. Fig. 33a, b, c. Fine-grained basalt, generally glassy pillow or flow fragments. Many are very vesicular (to 10%). Numerous textural variations. Fe-Mn coatings to 1 cm. Fig. 34. HECK SEAMOUNT CHAIN 7 East Peak Cryptocrystalline clinker, irregular pieces of flows, glass fragments, and one piece of holocrystalline, intergranular basalt. Fig. 35a, b, c, d. 8 Heck Seamount (north) 9 Heck Seamount (south) 10 Heck Seamount (west) 11 West Peak HECKLE SEAMOUNT CHAIN 12 East end south flank 13 West Peak Irregular fragments of cryptocrystalline and glassy basalt. Fig. 3Se, f. Aquagene tuff, chips of basalt and glass. Fig. 43d. e. f. Fig. 44a, b. Irregular fragments of fine-grained and glassy basalt, and tuffaceous material. Fig. 36a. Fine-grained, generally cryptocrystalline and glassy basalt, commonly with thick (to 6 mm) accumulations of Fe-Mn oxides. Fig. 36b. Fine-grained (intersertal) basalt. Pillow fragments with wide glass selvages and Fe-Mn coatings. Fig. 36c, d. Fine-grained (intersertal) basalt with thin glass rim and Fe-Mn coating to 1 cm. Fig. 36e. 103 Many of the basalt fragments have black glass selvages of various thicknesses, usually with a thin outer orange-brown layer of hydrated glass (palagonite). Textural variations occur between the glass selvages and the crystalline interiors. In thin section the glass inside the palagonitized zone is observed to be pale brown, isotropic sideromelane, commorily containing microphenocrysts of plagioclase, clinopyroxene, and/or olivine (Fig. 33a) and in some specimens cracks caused by contraction during cooling (Fig. 35f) . Vescicles, cracks, and microphenocrysts in the sideromelane of many specimens are rimmed by incipient, fibrous brown crystal growths, and the inner zone of clear glass is speckled with dark brown varioles, which generally nucleate around microphenocrysts and vesicles (Fig. 33b). Under high magnification, these varioles are observed to be composed of fibrous or feathery microlites dusted with minute opaque minerals, which give them a fuzzy appearance. The varioles increase in density and size away from the margin until they form a nearly opaque, dark brown variolitic zone (Fig. 33b). The cryptocrystalline nature of the minerals comprising the varioles prohibits optical investigation, but tracing the progressive growth of the varioles into the aphanitic interiors of the basalts suggests that clinopyroxene is the dominant mineral in the varioles, with subsidiary opaque minerals, plagio-clase, and glass. The groundmass in the variolitic zone is referred to as tachylite (turbid glass), as opposed to sideromelane (Fuller, 1932). Towards the interiors of the basalt fragments, the opaques segregate around fan-shaped and plumose aggregates (Fig. 33c), which may eventually be observed to be broken up into discrete clinopyroxene crystals among plagio-clase laths (Fig. 36d). The interior groundmass textures vary from hyalopilitic, in which microlites of plagioclase lie in a dark brown variolitic (tachylitic) 104 groundmass, to intersertal, in which tachylite and small crystals f i l l the interstices among plagioclase laths, to intergranular, in which interstices between plagioclase laths are occupied by granules of clinopyroxene and opaque minerals. Holocrystalline specimens generally have intergranular textures. The variations in texture observed in the basalts from the study area are ascribed to differences in cooling rates of the lavas, and, hence, in position in a pillow or flow relative to the exterior and in the size of the pillow or flow. Marshall and Cox (1971) present a thorough discussion of this topic. Juan de Fuca Ridge Basalts from Juan de Fuca Ridge include pillows and pillow fragments, fragments of flows, and angular massive blocks. The pillows and pillow fragments have glass selvages (Fig. 31c, d, a). These specimens have microcrystalline groundmasses even in their interiors, with interstitial cryptocrystalline minerals or glass, and ground-mass textures range from variolitic to hyalopilitic (Fig. 31b) or inter-sertal. Many of the pillow fragments from Station 2 have rough coatings of palagonite breccia (Fig. 30d). Some of the fragments identified as probable pieces of flows also have glass selvages, but they lack the characteristic pillow fragment shape. Most specimens lack glass selvages and are of irregular shapes (Fig. 30a). However, al l specimens have microcrystalline groundmasses throughout, with interstitial cryptocrystalline material and tachylite, and textures are hyalopilitic (Fig. 30b) or intersertal (Fig. 30f) . FIGURE 30. Basalts from Juan de Fuca Ridge (Middle Ridge) a. Hand specimen of glomeroporphyritic plagioclase basalt. Specimen 1-7. b. Photomicrograph of glomeroporphyritic basalt showing plagioclase xenocrysts (containing stringers of tachylite) and olivine pseudomorphs in hyalopilitic groundmass. Specimen 1-5. Plane light X36. c. Photomicrograph of glomeroporphyritic basalt showing tachylite inclusions in plagioclase xenocrysts. Specimen 1-8. Plane light X30. d. Massive, holocrystalline basalt from Station 2 (3 specimens on right) contrasted with glassy, breccia-encased basalt from same station. e. Photomicrograph of holocrystalline basalt, showing plagioclase, clinopyroxene, and opaque minerals in intergranular texture. Specimen 2-1. Plane light X125. f. Photomicrograph of glassy basalt showing plagioclase laths, opaque minerals, and cryptocrystalline clinopyroxene and palgioclase forming an intersertal texture. Irregular white areas are vesicles. Specimen 2-2. Plane light X125. 105 FIGURE 30. Basalts from Juan de Fuca Ridge (Middle Ridge). FIGURE 31. Basalts from Juan de Fuca Ridge (West Ridge and Endeavour Seamount) a. Hand specimen of glomeroporphyritic olivine basalt with typical pillow fragment shape and wide glass selvage. Specimen 3-3. b. Photomicrograph of glomeroporphyritic basalt, showing olivine xenocrysts in hyalopilitic groundmass. Specimen 3-3. Plane light X36. c. Two pillow fragments of glomeroporphyritic plagioclase basalt, fitted together at broken neck to show original configuration. An older break faces viewer. Specimen 4-1. d. Same as c , but showing more clearly where palagonitized glass has spalled off to expose underlying fresh glass, dotted with plagioclase xenocrysts. Specimen 4-1. e. Fresh plagioclase xenocryst in cryptocrystalline basalt. Specimen 4-1. 106 FIGURE 31. Basalts from Juan de Fuca Ridge (West Ridge and Endeavour Seamount). 107 The angular massive blocks of basalt (Fig. 30d) are holocrystalline and coarser in grain size than the pillows, pillow fragments, and flow fragments. They have intergranular textures (Fig. 30e). Such blocks may be from the interiors of thick flows, or they may represent intrusive rocks. They were recovered only from Station 2 on Middle Ridge, near the base of a fault scarp (Profile F-F'--Plate VI) which may account for this exposure of rocks apparently from deeper levels. These holocrystalline basalts as well as the fragments of pillows and flows collected at Station 2 are non-porphyritic, but a l l the specimens from the other three stations on Juan de Fuca Ridge are porphyritic. The fragments of flows and the pillow from Stations 1 and 4 respectively contain large crystals of calcic plagioclase (Fig. 30a, 31c, d, e), whereas the pillow fragments from Station 3 contain forsteritic olivine (Fig. 31a). These plagioclase and olivine crystals generally show strong effects of resorption in the form of rounded and corroded outlines. The crystals tend to occur in clusters (glomeroporphyritic). The plagioclase xenocrysts commonly contain inclusions of tachylite (Fig. 30b, c). The origins of these large crystals of plagioclase and olivine are discussed on pages 115 to 117. The pillow basalts from Station 3 and 4 are very fresh and no altera-tion products were observed in hand specimen or thin section, with the exception of palagonite on the exterior glass. In contrast, the specimens from Stations 1 and 2 are altered. No fresh olivine was observed in their thin sections. Many specimens have dark alteration rims, and alteration products occur in interstices and vesicles (see p. 149-154 for a discussion of the alteration of the basalts). c 108 Sovanco Ridge Basalts having a wide range of form and texture were dredged from the two stations on Sovanco Ridge. Angular blocks of holocrystalline basalt (Fig. 32a) were collected from near the base of the fault scarp forming the eastern end of the ridge (Profile G-H--Plate VII). These specimens have intergranular textures, and include fine-grained (Fig. 32b) and medium-grained (Fig. 32c) types. They may represent either the interiors of thick flows or intrusive rocks, like the holocrystalline specimens from Station 2 on Juan de Fuca Ridge. As at Station 2, faulting may account for the exposure of basalts apparently from deeper levels. Flow fragments, in some cases with glass selvages, of glomero-porphyritic plagioclase basalt (Fig. 32d) were also collected from Station 5. These specimens are similar to those dredged from Station 1 on Juan de Fuca Ridge. The large plagioclase crystals are calcic in composition (An 85) but commonly have outer zones of more sodic composition. They generally contain inclusions of tachylite, have rounded or corroded outlines (Fig. 32e, f ) , and are interpreted to be xenocrysts (see discussion, p.115-117). Groundmass textures in the interior parts of these specimens are intersertal (Fig. 32e, f ) . Fragments of non-porphyritic basalt, many with glass selvages, were also obtained at Station 5. Some are pillow fragments; most are probably fragments of flows. Glassy specimens show the typical transition from sideromelane to hyalopilitic or intersertal interiors (Fig. 33a, b, c). The specimens from the other station on Sovanco Ridge (Profile L-L1--Fig. 25) include pillow fragments (Fig. 34a), platy flow fragments, and irregular pieces encased in palagonite breccia (Fig. 34b). Textures in the interiors of fragments are generally intersertal (Fig. 34c, d, e) or hyalopilitic (Fig. 34f). The presence of fresh groundmass olivine crystals FIGURE 32. Basalts from the eastern end of Sovanco Ridge a. Hand specimens (broken, lower right, and cut surfaces) of holocrystalline basalt from Station 5 showing dark and yellow-brown tinted alteration rims. b. Photomicrograph of holocrystalline basalt (fine-grained subgroup), showing plagioclase laths, small clinopyroxene crystals, scattered opaque minerals, and alteration products forming an intergranular texture. Specimen 5-9. Plane light X36. c. Photomicrograph of holocrystalline basalt (medium-grained subgroup), showing plagioclase laths, clinopyroxene crystals, opaque minerals, and alteration products forming an intergranular texture. Specimen 5-8. Plane light X36. d. Hand specimen of altered glomeroporphyritic plagioclase basalt with wide glass selvage partly altered to yellow palagonite. Specimen 5-6. e. Photomicrograph of glomeroporphyritic basalt, showing plagioclase xenocrysts containing stringers of tachylite, Groundmass texture is intersertal. Specimen 5-10. Plane light X36. f. Same as e., crossed nicols. To show outermost sodic rims on plagioclase xenocrysts. 109 FIGURE 32. Basalts from the eastern end of Sovanco Ridge. FIGURE 33. Photomicrographs of basalt from the eastern end of Sovanco Ridge, illustrating the typical transition from glass selvage to crystalline interior observed in basalts from the study area. a. Sideromelane with plagioclase and olivine microphenocrysts and incipient fibrous crystallization around vesicles and along fractures. Specimen 5-11. Plane light X36. b. Transition between sideromelane and variolitic zone, showing nucleation of varioles around microphenocrysts and vesicles. Specimen 5-11. Plane light X36. c. Plagioclase phenocrysts and partially altered olivine phenocrysts in hyalopilitic interior. Groundmass is tachylite. Specimen 5-11. Plane light X36. 110 FIGURE 33. Photomicrographs of basalt from the eastern end of Sovanco Ridge, illustrating the typical transition from glass selvage to crystalline interior observed in basalts from the study area. FIGURE 34. Basalts from the southern flank of a peak on Sovanco Ridge a. Hand specimen showing typical pillow fragment shape with radial fractures. The banded "oolitic" texture is due to radiating clusters of tiny crystals (palgioclase?). Specimen 6-4. b. Angular basalt fragment encased in palagonite breccia. Specimen 6-3. c. Photomicrograph of variolitic to intersertal basalt containing fresh olivine phenocrysts. Specimen 6-5. Plane light X120. d. Photomicrograph of basalt with intersertal texture. Specimen 6-2. Plane light X36. e. Photomicrograph of basalt with intersertal texture at greater magnification than d. Specimen contains higher proportion of opaque minerals than specimen 6-2. Specimen 6-6. Plane light X125. f. Photomicrograph of basalt with hyalopilitic texture, containing plagioclase and olivine crystals and about 6% opaque minerals. Specimen 6-1. Plane light X36. FIGURE 34. Basalts from the southern flank of a peak on Sovanco Ridge. 112 is characteristic of many specimens (Fig. 34f). The basalts from Station 6 are the only specimens dredged from the study area with significant amounts of fresh groundmass olivines. Small, generally euhedral phenocrysts of olivine are also common (Fig. 34c). Heck and Heckle Seamount Chains Small, irregular fragments of basalt having a variety of forms, many reminiscent of clinker, aa, and pahoehoe lavas, were commonly dredged from the seamount chains (Fig. 35a, b; Fig. 36b). However, pieces of more massive basalt (Fig. 35c, 36e) and fragments of pillows (Fig. 36c) are also common. Some large pieces of basalt are exceptionally glassy (Fig. 35e). In general, much greater quantities of glass were recovered in the seamount chain dredge hauls than in hauls from either Juan de Fuca or Sovanco Ridges. The seamount chain basalts contain scattered plagioclase and, more rarely, olivine phenocrysts (Fig. 35b, c; Fig. 36c). Equivalents of the highly porphyritic basalts so common on Juan de Fuca and Sovanco Ridges were not dredged from the seamount chains. Furthermore, the scattered phenocrysts in the seamount chain basalts are apparently less calcic and forsteritic than the large plagioclase and olivine crystals in the porphyritic ridge basalts. The interiors of most specimens have intergranular textures, but these show considerable variation in detail (Fig. 35d; Fig. 36a, d). Specimens from Stations 11, 12, and 13 (those furthest from Juan de Fuca Ridge along the axes of the seamount chains) generally show effects of altera-tion and have thick deposits of ferro-manganese oxides (Fig. 36b, c, e). Tuffaceous deposits were commonly dredged from the seamount chains (see p. 155). FIGURE 35. Basalts from Heck Seamount Chain a. Specimens of clinker from Station 7. b. Broken surface of cryptocrystalline clinker, showing altered rim, plagioclase and olivine phenocrysts, and a large vesicle. Specimen 7-4. c. Cut surface of holocrystalline basalt, showing dark alteration rim and scattered plagioclase phenocrysts. Specimen 7-1. d. Photomicrograph of holocrystalline basalt, showing palgioclase, clinopyroxene, and opaque minerals in an intergranular texture. Specimen 7-1. Plane light X125. e. Glassy basalt from Station 8. Outer glass is altered to yellow palagonite. f. Photomicrograph of sideromelane, cut by polygonal fractures probably caused by contraction during cooling, and containing variolitic crystallization. Specimen 8-4. Plane light X110. 113 FIGURE 35. Basalts from Heck Seamount Chain FIGURE 36. Basalts from Heck and Heckle Seamount Chains a. Photomicrograph of massive basalt from Station 10, showing plagioclase laths, clinopyroxene crystals, and opaque minerals in intergranular texture. Specimen 10-1. Plane light XI25. b. Basalt specimen from Station 11 with ferro-manganese oxide coating. c. Pillow fragment of fine-grained basalt with scattered plagioclase phenocrysts. Specimen has irregular ferro-manganese oxide coating, glass selvage, and extensive dark altered areas. Specimen 12-5. d. Photomicrograph of fine-grained basalt from Station 12, with plagioclase, clinopyroxene, and opaque minerals in intersertal to intergranular texture. Specimen 12-1. Plane light X125. e. Fine-grained basalt with thick ferro-manganese oxide coating, dark alteration rim, and fresh grey core (emphasized by plastic spray coating). Specimen 13-1. 114 FIGURE 36. Basalts from Heck and Heckle Seamount Chains. 115 Significance of the Petrography The basalts from the study area show considerable variation in external form, texture, degree of crystallinity, and extent of alteration, such as is commonly observed among submarine basalts. Some of the variations obviously do not have regional implications; for example, the textural varia-tion from the chilled glass selvage through the variolitic zone to the crystalline interior which is a function of cooling rate. However, some of the variations may be interpreted in terms of significant differences in the histories of the magmas or in environments during and after eruption. Such differences are the topic of this section. Porphyritic basalts Highly porphyritic basalts were dredged only from Juan de Fuca and Sovanco Ridges. The basalts from the seamount chains contain only very scattered small plagioclase and olivine phenocrysts of less calcic and forsteritic compositions, respectively, than those in the porphyritic ridge basalts. The large plagioclase crystals in the porphyritic basalts from Stations 1, 4, and 5 have several features in common. They are generally equant in form, up to 1 cm in diameter, and in glomeroporphyritic clusters. Many crystals have embayed edges and rounded outlines, and inclusions of tachylite are common. Their central zones are very calcic (generally greater than An85), but most have one or more outer zones of more sodic compositions. Many of the forsteritic olivine crystals in specimens from Station 3 also have rounded outlines and tend to occur in clusters. Large crystals of plagioclase, and more rarely of olivine, of similar characteristics have been described in many basalts from the 116 Mid-Atlantic Ridge. On the basis of their rounded and corroded outlines, most workers have called them xenocrysts (Muir and Tilley, 1964; Aumento, 1968; Melson and Thompson, 1971). However, Miyashiro et_ al_ (1969) did not consider that the evidence for xenocrystic origin to be decisive, and regarded the crystals in their specimens as true phenocrysts. The question is of considerable importance with regard to evaluation of the bulk chemical analyses of porphyritic specimens. Kuno (1950) described plagioclase crystals from Hakone Volcano in Japan with similar characteristics which he ascribed to partial remelting of xenocrystic plagioclase, particularly along cleavages. The melted fraction forms irregular or vermicular shapes which may be connected by narrow channels and may solidify to form honey-combed structure within the plagioclase crystal. Remelting occurred when the xenocrysts were in a magma in equili-brium with more calcic plagioclase. However, Drever and Johnston (1957) demonstrated that, in the case of olivine crystals at least, large phenocrysts can grow "in situ". Further-more such crystals may be embayed, corroded, or rounded in appearance, presumably as a result of environmental factors which are not yet understood. They may include glass as a result of enclosure or partial enclosure of groundmass during crystal growth. However, Drever and Johnston noted that most workers attribute such inclusions to partial melting. It is readily conceivable that "true" phenocrysts could acquire xenocrystic characteristics i f they grew- in a magma chamber at intermediate depth and were subsequently moved rapidly with their magma into lower pressure regions. Temporary superheating would cause the observed resorption features. However, i f such early formed phenocrysts underwent processes of crystal accumulation, by gravitational sinking or floating, or perhaps as a result of 117 convection in the magma chamber, then they could be considered to be "cognate" xenocrysts. Such crystals could develop the observed resorption features either by movement into deeper, hotter portions of the magma chamber, or by rapid movement to higher levels subsequent to accumulation. This is the origin postulated by Aumento (1968) and Melson and Thompson (1971) Hence, the origins of the large crystals of plagioclase and olivine on the porphyritic ridge basalts are not certain on the basis of petrography alone. However, the petrographic evidence combined with the chemical results suggests that the crystals are indeed xenocrysts, because the compositions of the porphyritic basalts after the removal of approximate modal concentrations of xenocrysts are very similar to the compositions of the non-porphyritic ridge basalts from the study area, and to ocean ridge basalts in general. If the crystals are true phenocrysts, then the original magmas were anomalously rich in Al and Ca (in the case of plagioclase phenocrysts) or Mg (in the case of olivine phenocrysts). The chemistry is further discussed on pages 124 to 126. The reason for the glomeroporphyritic characteristic of these basalts is not known. If the sampling is assumed to be representative, then the lack of porphyritic specimens from the seamount chains indicates that magmas do not reside in intermediate depth magma chambers below those features. Magnetic and petrographic evidence suggests that the seamounts of the Heck and Heckle Chains grew to virtually their final sizes while s t i l l at or near Juan de Fuca Ridge crest. This would require large amounts of magma in a short time span, so i t is reasonable that the magmas did indeed have very brief residences in the magma chambers. 118 Faulting Although pillow fragments were dredged from the ridges as well as the seamounts, the massive, smooth-sided blocks of holocrystalline basalt were recovered only from the ridges. Even there, they were dredged only from areas where the seismic reflection profiles indicate that major faulting of the igneous basement has occurred. Hence, these faults, which are very unlikely to occur on seamounts, expose deeper portions of flows, or perhaps dykes. They also result in steep slopes and talus deposits, where hetero-geneous collections of basalt can be dredged. Mode of Eruption The basalt fragments from the seamount chains tend to be more irregular in form than the ridge basalts, and although usually glassy, many are not pillow fragments. Many pieces have forms suggestive of fragments of clinker, or aa and pahoehoe flows. Although a few samples of hyaloclastites (interpillow breccias) were recovered from the ridges, hyaloclastites and tuffaceous deposits of many types were collected from the seamount chains (see p. 155-160). The impression is that the seamount chain eruptions tend to be more violent in nature than the ridge eruptions. This is consistent with the difference between central vent eruptions, where evidence is commonly found of extensive lava-water interaction during the eruption, and fissure eruptions, where "quiet" deep-sea effusion occurs (Bonatti, 1970). The shallower depths of many seamount chain eruptions may also account for some of the differences. 119 Chemistry of the Basalts The fresh basalts from Juan de Fuca Ridge, Sovanco Ridge, and the seamount chains in the study area are classified as low-potassium tholeiites of the ocean ridge type on the basis of major oxide compositions (Tables III and IV). However, detailed examination of the results reveals minor variations among the analyses. As demonstrated in Appendix B, Part 2, concerning the accuracy and precision of the chemical analyses, i t is reasonable to consider the reproducible small variations to be significant. The totals for the chemical analyses of the fresh basalts range from 94.27 to 99.58. The low totals are partly explained by the fact that the water contents of the samples were not determined. These determinations were attempted but reproducible results were not obtained. p2®5> ^2> anc* the other more minor constituents of basalts were also not determined. Furthermore, because the ferric/ferrous iron ratios were not determined, total iron is given as FeO, whereas a certain proportion of the iron is undoubtedly present in the basalts as Fe20j. These factors together can easily account for 2 or 3% on the basis of comparisons with analyses of similar ocean ridge basalts from other areas for which these constituents have been determined. However, part of the discrepancy between the totals of Tables III and IV and 100% is caused by the use of a rapid analytical technique (see Appendix B, Part 2). Both glass selvages and crystalline interiors of four specimens were analyzed for comparison, because of the possible changes brought about during the i n i t i a l reaction of the molten lava with sea water. In addition, to gain some idea of subsequent changes, samples from both the dark alteration rims and the lighter cores of six specimens were analyzed, as well as samples TABLE III. Partial chemical analyses and normative compositions (Wt. %) of fresh basalts from Juan de Fuca Ridge (Col. 1-10) and Sovanco Ridge (Col. 11-17). Norms calculated using Fe3+/Fe2+ = 0.1 S. I. =, Solidification Index = MgO x 100/MgO + FeO + Na20 + K20 (after Kuno, 1968) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Specimen ^  ^ Number 1-2 1-3 1-4 1-5 2-1 2-2 2-3 3-1 4-1 5-1 5-2 5-3 5-4 5-5 6-1 6-2 Si02 47.4 48 5 48. 2 48.2 46.2 49.6 52.1 50.2 47.7 49.1 48.6 47.6 47.9 49.5 48.5 47.1 50.1 A1203 19.0 15 8 18. 4 19.2 16.4 15.1 14.4 13.8 15.6 15.7 19.6 18.4 15.2 15.2 16.2 16.S 16.5 Ti02 0.83 1 03 0. 91 0.82 1.02 1.17 2.08 '.1.98 0.75 0.88 0.87 0.76 1.31 1.16 1.20 1.55 0.99 FeOy 7.3 8 6 8. 0 7.2 8.9 8.9 9.3 9.0 8.1 8.3 7.3 7.3 9.1 8.8 8.3 8.5 7.3 MgO 6.4 6 8 7. 1 6.6 10.0 7.5 7.5 6.6 11.8 7.6 6.5 6.1 7.7 7.3 7.2 7.8 8.1 CaO 12.8 12 4 12.3 12.4 11.1 11.2 10.6 10.6 10.7 12.0 12.2 11.9 11.0 11.1 11.3 10.1 11.5 Na20 1.9 2 0 1. 9 2.0 2.4 2.2 3.2 2.3 2.0 3.6 2.0 2.0 2.7 2.1 2.6 2.5 2.3 K20 0.08 0.14 0. 10 0.09 0.13 0.05 0.19 0.25 0.04 0.04 0.08 0.08 0.05 0.06 0.11 0.42 0.03 MnO 0.13 0 15 0. 12 0.13 0.17 0.15 0.21 0.21 0.15 0.16 0.12 0.13 0.16 0.16 0.16 0.13 0.15 Total 95.84 95 42 97. 03 96.64 96.32 95.97 99.58 94.94 96.84 97.38 97.27 94.27 95.12 95.38 95.57 94.60 96.97 S.I. 40.05 38 90 41. 60 41.52 46.70 40.31 36.93 36.27 53.76 38.56 40.97 39.24 39.35 39.93 .39.64 40.56 45.64 Q - 0 1 - - 1.5 - 3.6 - - - 0.3 - 2.0 - - 0.4 Ne - - • 0.2 - - - - 2.8 - - - - - - -Or 0.47 0 83 0. 59 0.53 0.77 0.30 1.12 1.50 0.24 0.24 0.47 0.47 0.30 0.35 0.65 2.48 0.18 Ab 15.7 16 9 16. 4 17.2 20.1 18.2 27.2 19.8 17.1 25.0 16.8 16.8 22.7 17.6 21.8 21.0 19.0 An 43.3 33 6 41. 2 43.0 33.6 31.3 24.4 26.5 33.3 26.8 44.4 41.1 29.2 31.9 32.4 32.7 34.8 Di 16.3 22 8 16. 3 15.0 17.4 19.6 23.0 21.3 15.7 26.8 13.0 14.5 20.5 18.8 19.4 14.2 18.0 Hy 14.8 17 9 16. 1 15.6 - 21.2 17.7 17.4 12.9 19.3 18.4 10.5 21.1 12.8 11.2 21.5 01 2.5 3. 6 2.7 21.1 - 0.88 •- 14.9 12.9 0.6 - 7.9 -' S.2 8.9 -Mt 1.2 1 4 1. 3 1.2 1.4 1.4 1.5 1.5 1.3 1.3 1.2 1.'2 1.5 1.4 1.4 1.4 1.2 11 1.6 2 0 1. 7 1.5 1.9 2.2 4.0 3.8 1.4 1.7 1.7 1.4 2.5 2.2 2.3 2.9 1.9 FeOr = Total Fe as FeO 121 TABLE IV. Partial chemical analyses and normative compositions (Wt. %) of fresh basalts from Heck (Col. 1-6) and Heckle (Col. 7-10) Seamount Chains. Norms calculated using Fe3+/Fe2+ =0.1. S. I. = Solidification Index = MgO x 100/MgO + FeO + Na20 + K20 (after Kuno, 1968) 1 2 3 4 5 6 7 8 9 10 Specimen y ^ Number 8-1 8-2 8-3 10-1 11-1 12-1 12-2 12-3 13-1 Si02 A1203 Ti02 FeOr MgO CaO Na20 K20 MnO Total S. I. 50.8 50.7 48.7 49.0 50.9 50.2 50.5 49.0 49.5 50.1 15.1 15.0 14.7 14.9 14.8 14.6 14.6 15.7 14.7 14.4 1.00 0.87 0.88 0.84 0.98 0.94 1.22 1.33 1.22 0.97 8.4 8.8 9.1 8.9 9.6 8.5 8.3 8.0 8.2 7.7 7.6 8.2 7.4 6.8 7.8 7.7 7.0 7.0 7.2 7.1 12.5 11.8 12.3 12.2 11.8 12.5 11.6 11.6 11.5 11.8 2.1 2.6 1.7 2.5 2.4 2.3 2.1 2.6 2.5 2.1 0.05 0.04 0.08 0.06 0.09 0.04 0.08 0.11 0.09 0.12 0.16 0.17 0.17 0.16 0.18 0.16 0.18 0.18 0.18 0.17 97.71 98.18 95.03 95.36 98.55 96.94 95.58 95.52 95.09 94.46 41.80 41.82 40.37 37.12 39.06 42.03 40.29 39.40 39.96 41.55 1.4 1.3 _ _ _ 3.6 . 0.3 2.9 Q Ne Or Ab An Di Hy 01 Mt II 0.30 17.9 31.5 2S.0 18.5 1.4 1.9 0.24 21.8 29.2 24.0 16.2 3.9 1.4 1.7 0.47 14.6 32.1 23.8 19.7 1.5 1.7 0.35 0.53 0.24 0.47 0.65 0.53 0.71 21.5 20.5 19.7 17.6 22.3 21.2 18.1 29.1 29.2 29.2 30.1 30.8 28.5 29.2 25.6 24.2 26.9 22.4 21.7 23.4 23.8 12.2 19.9 17.7 17.8 13.7 17.6 16.7 3.5 1.1 0.2 - 2.8 - -1.4 1.6 1.4 1.3 1.3 1.3 1.3 1.6 1.9 1.8 2.3 2.5 . 2.3 1.8 FeOj = Total Fe as FeO 122 from four obviously altered rocks. These data, together with the results of the analysis of glass specimen 9-1, are presented and discussed in separate sections following the presentation of the data for fresh and crystalline basalt. A brief description of each specimen analyzed is given in Appendix B, Table XIX. The solidification index (SI = MgO X 100/MgO + FeOT + Na20 + K20) has been calculated for each analyzed specimen, and is given in Tables III and IV. The solidification index is the percentage of MgO in the triangular diagram MgO - FeO-r - Na20 + K20 (MFA diagram) . In any fractionation trend plotted on an MFA diagram, composition of magma changes so as to decrease the SI value as fractionation proceeds. It is assumed that the SI value decreases proportionately with the amount of residual liquid (Kuno, 1968). Kuno suggested that undifferentiated magmas have solidification indices between 40 and 35, and that rocks with SI higher than about 40 appear to be products of accumulation of olivine crystals. SI is also used to construct variation diagrams rather than Si02 because the silica-variation diagram has the dis-advantage of emphasizing variation in the late stage of fractionation more than that in the early and middle stages (Kuno, 1968). A computer program made available by P. B. Read was used for the calculations of the CIPW normative compositions. Because of the absence of data about the oxidation state of iron and to enable meaningful comparisons of normative compositions, the ferric/ferrous iron ratios were set at 0.1 for the calculations (after Bass, 1971). This ratio is virtually the lowest observed in basalts. The distinctive characteristics of the chemical compositions of the basalts from each of Juan de Fuca Ridge, Sovanco Ridge, and Heck and Heckle 123 Seamount Chains are discussed separately; a general discussion of the signi-ficance of the results follows the detailed picture. Each analyzed specimen is identified by its dredge station number followed by a sample number (for example, 4-1). Juan de Fuca Ridge (Table III) The specimens analyzed from Juan de Fuca Ridge f a l l into three distinct chemical types: (a) Basalt characterized by low potash, titania, and soda (specimen 2-1 and porphyritic specimens from Stations 1 and 3) (Type A) (b) Basalt characterized by low potash, low titania, and high soda (specimen 4-1) (c) Basalt characterized by high potash and high titania (specimens 2-2 and 2-3) The differences among these three types are not great, the number of analyzed samples is small, and the other oxides within any one type in some cases show considerable variation, whereas some oxides show similarity from one type to another. Nevertheless, on the basis of available data, the distinctions are considered to be valid. (a) Basalt characterized by low potash, titania, and soda (Type A) These characteristics are common to the majority of fresh basalts analyzed from the study area, including those from the seamount chains and most of those from Sovanco Ridge and Juan de Fuca Ridge. Those from Sovanco and Juan de Fuca Ridges are hereafter referred to as Type A basalts, to distinguish them from variant basalts from those locations. The holocrystalline specimen 2-1 from Station 2 on Middle Ridge is representative of this type. 124 The porphyritic natures of the specimens from Stations 1 and 3 are reflected in their chemical compositions. However, comparisons of oxides not involved in phenocrysts (xenocrysts) indicate that the porphyritic basalts from those stations are probably not chemically distinct from the non-porphyritic specimen 2-1. The porphyritic basalts from Station 1 on Middle Ridge have rela-tively large alumina and lime contents (Table III, Col. 1-5). If the large plagioclase crystals in these basalts are considered to be xenocrysts as discussed previously (p. 115) and the basalts to be partial cumulates, then . the effective compositions of the xenocrysts can be removed from the whole rock analyses as described by Aumento (1968) for similar porphyritic basalts from the Mid-Atlantic Ridge near 45°N. Four of the analyses of Station 1 basalts were recalculated by removing modal concentrations of plagioclase xenocrysts of composition An85 (Table V, Col. 1-4). Both alumina and lime are reduced by the removal of the xenocrysts, whereas the other oxides show small increases. As demonstrated on plots of lime vs. alumina (Fig. 41b, p. 141) and iron vs. magnesia (Fig. 41a) the effect is to move the Station 1 points closer to those of non-porphyritic Juan de Fuca (and Sovanco) Ridge basalts. Specimen 1-2 (Table III) contains only 3% xenocrysts, and its chemical analysis is similar to the recalculated analyses. Similarly, the high magnesia content of specimen 3-1 from West Ridge reflects the presence of about 10% crystals of forsteritic olivine, and the chemical analysis was recalculated to remove the modal concentration of olivine xenocrysts (Table V, Col. 8). The high alumina and lime contents suggest that some of the plagioclase phenocrysts in this specimen may also be xenocrysts. TABLE V. Recalculated partial chemical analyses of porphyritic basalts from Juan de Fuca Ridge (Col. 1-5, and 8) and Sovanco Ridge (Col. 6 and 7). Columns 1-7: remove plagioclase of composition An85. Column 8: remove olivine of composition Fo90. Original compositions are given in Table III 1 2 3 4 5 6 7 8 Specimen Number 1-1 1-3 1-4 1-5 4-1 5-1 5-2 3-1 Si02 47.7 48.5 48.6 46.3 49.4 49.0 47.9 48.6 A1203 16.5 15.7 16.6 14.5 13.7 17.1 16.8 17.4 Ti02 0.98 1.08 0.97 1.14 0.98 1.03 0.85 0.84 FeOj 8.7 9.5 8.5 9.9 9.3 8.6 8.2 8.2 MgO 7.6 8.4 7.8 11.2 8.5 7.7 6.8 7.4 CaO 12.1 11.5 11.6 10.5 11.5 11.3 11.4 11.9 Na20 1.9 2.0 2.1 2.5 3.8 2.0 2.0 2.2 K20 0.09 0.12 0.11 0.15 0.05 0.09 0.09 0.05 MnO 0.15 0.14 0.15 0.19 0.18 0.14 0.15 0.17 Total 95.72 96.94 96.43 96.38 97.41 96.96 94.19 96.76 Crystals Removed % 15 15 15 10 10 15 10 10 Type Plag. Plag. Plag. Plag. Plag. Plag. Plag. 01. Fe0T = Total Fe as FeO 126 The computed analyses are likely to have large errors owing to the uncertainties in determining actual xenocryst compositions and concentrations, and to the lack of consideration of the effects of resorption. However, the mean of seven analyses of porphyritic basalts, including six recalculated analyses of specimens from Stations 1 (excluding anomalous specimen 1-5), 3, and 5, as well as the analysis of the sparsely porphyritic specimen 1-2, has been calculated, and the standard deviations of the oxides are not excessive (Table VI, Col. 2, p. 132). A comparison of this mean with the mean of four analyses of non-porphyritic Type A basalts from Stations 2 and 5 shows reasonable similarity, although soda and titania are somewhat discordant (Table VI). Hence, i t is thought that the porphyritic basalts represent essentially Type A magmas modified by crystal accumulation. The variations among the analyses from Station 1 are partly related to varying concentrations of "xenocrysts". However, more important is the fact that a l l specimens analysed show some effects of alteration, and, although the freshest portions were sampled for analysis, the degree of "freshness" undoubtedly varied. The unusually high magnesia content of specimen 1-5 may reflect more olivine, as indicated by the normative composition, although this was not evident in hand specimen or thin section. (b) Basalt characterized by low potash, low titania, and high soda Because the analysis reported for the one representative of this type (specimen 4-1) is the mean of two duplicate analyses (both of which are presented in Appendix B with the discussion of precision of the chemical results), the high soda content is thought to be correct. 127 With the exception of the high soda content, the composition of specimen 4-1 is very similar to that of the Type A basalt. The recalculated analysis after removal of the modal concentration of plagioclase xenocrysts is given in Table V. The high soda content (3.8%) of the basalt is unusual in view of the low potash content (0.05%), and results in the presence of nepheline in the normative composition. For comparison the average ocean alkali basalt (Engel et a l , 1965) contains 3.99% Na20, but 1.66% K20. The other oxides are also more like ocean ridge tholeiites than alkali basalts, and on Figure 40b (p. 139) specimen 4-1 plots with the other basalts from the study area in the ocean ridge basalt fie l d . Hence the high soda content of specimen 4-1 is an enigma. Spilitized basalts of similar high-soda, low-potash compositions have been described (Cann and Vine, 1965; Cann, 1969) but specimen 4-1 appears to be fresh and contains unaltered calcic plagioclase xenocrysts. However, the glass rim of 4-1 contains even more soda than the interior (Table IX, Col. 3 and 4, p. 152), so one explanation might be unusual sea water contamination. (c) Basalt characterized by high potash and high titania The higher titania and potash contents, and the lower solidification indices (Kuno, 1968), of specimens 2-2 and 2-3 suggest that this basalt is slightly more differentiated than the other types. The titania contents of these two specimens are by far the highest determined in basalts from the study area. Correspondingly, in thin section these specimens show an unusually high content of opaque minerals, probably titano-magnetite (Fig. 30f). The potash contents of the two specimens are also unusually high, as is the soda content of specimen 2-2. The difference between the soda contents of the two specimens suggests that an error in one of the chemical analyses 128 is responsible, probably the low soda of 2-3 as differentiation would be expected to result in an increase in soda as well as potash and titania (Kuno, 1968), and the altered rim of specimen 2-2 also contains high soda (Table VIII, Col. 4, p. 151). The petrographic heterogeneity of basalts from Station 2 is indicative of chemical heterogeneity as well. However, i t will be seen that this is not true at Station 5 on Sovanco Ridge, where holocrystalline and glassy basalts are similar in composition. Sovanco Ridge (Table III) Three chemically different basalt types are recognized from Sovanco Ridge: (a) Basalt characterized by low potash, titania and soda (specimens 5-1, 5-2, 5-3, 5-4, 5-5) (Type A) (b) Basalt characterized by low potash, titania and iron but high magnesia (speciman 6-2) (c) Basalt characterized by high potash and high titania (specimen 6-1) Obviously the paucity of analyses makes the distinction of the latter two types very inconclusive. (a) Basalt characterized by low potash, titania, and soda (Type A) All the specimens analyzed from Station 5 on Sovanco Ridge possess these characteristics common to the majority of basalts from the study area, even although the analyses represent porphyritic (5-1, 5-2), holocrystalline (5-3, 5-4), and the fine-grained glassy (5-5) basalts. 129 The porphyritic basalts show the typically higher alumina and lime contents, reflecting the presence of the anorthitic plagioclase crystals which are considered to be xenocrysts as previously discussed. The composi-tions of the porphyritic rocks from Station 5 are similar to the compositions of those from Station 1. The effects of recalculation to remove modal concen-trations of xenocrysts are shown in Table V. These recalculated analyses, together with those from Stations 1 and 3, were used to calculate the mean presented in Table VI, Col. 2, p. 132). Both of the holocrystalline basalts analyzed represent the finer-grained sub-group (Fig. 32b). Their compositions are very similar to the composition of the holocrystalline basalt specimen 2-1 from Juan de Fuca Ridge. The fine-grained, glassy basalt specimen 5-5, despite its very different form and texture, is very similar in composition to the holocry-stalline basalts of Station 5, and hence to the Type A basalts of both Juan de Fuca and Sovanco Ridges. These three analyses were combined with the analysis of specimen 2-1 to give the mean non-porphyritic Type A ridge basalt analysis (Table VI, Col. 1). (b) Basalt characterized by low potash, titania, and iron, but high magnesia Specimen 6-2 is like the Type A basalt of Juan de Fuca Ridge and Station 5 on Sovanco Ridge in most oxide concentrations, but its total iron content is much lower and its magnesia content is higher. In thin section the specimen is observed to contain 3-5% euhedral olivine microphenocrysts which are not thought to be xenocrysts. Furthermore the low iron content could not be explained by the addition of olivine xenocrysts. More analyses are needed to determine the validity and importance of these variations. 130 (c) Basalt characterized by high potash and titania Specimen 6-1 is not the same as the high-potash, high-titania basalt analyzed from Juan de Fuca Ridge in that i t contains twice as much potash but much less titania than the specimens from Station 2. Other oxides, notably alumina and s i l i c a , are also very different. Like specimen 6-2, specimen 6-1 contains fresh olivine microphenocrysts, whose presence reflects the high magnesia content. Because of this high magnesia content, the solidification index of specimen 6-1 (Table III) is normal, and i t is unlikely that the composition of the specimen is the result of differentia-tion. However, specimen 6-1 is more weathered than the other basalts whose analyses are given in Table III. The total of its analysis is low, and may reflect large contents of water and F^Oj as a result of weathering. Increased I^O.is also a common effect of submarine weathering (see p. 153). Consequently whether or not specimen 6-1 differs significantly from specimen 6-2 is not known. Seamount Chains (Table IV) The chemical analyses of fresh basalts from the seamount chains show striking similarity. The average of ten analyses from the six seamount stations with the standard deviation for each oxide is shown in Table VI (p. 132). These basalts are low potassium tholeiites of the ocean ridge type and do not show any affinities with the alkali or transitional-to-alkali basalts frequently dredged from seamounts. The average analysis does not differ markedly from the mean analysis of the common Type A ridge basalt (Table VI, Col. 1 and 2). 131 Minor but consistent differences seem to exist between Heck and Heckle Chain basalts in the lower iron and magnesia contents of the Heckle basalts (specimens 12-1, 12-2, 12-3, 13-1). Also, the basalts from the eastern end of the Heckle Chain (Station 12) contain about 0.3% more titania than the other specimens. The three fine-grained and glassy basalt specimens analyzed from Station 12 are very similar to one another, and the variations observed in their chemical compositions are considered to represent variations inherent in the analytical technique. The altered basalts from the seamount chains, as well as the glass margins of some specimens, show a high-alumina trend which distinguishes them from the fresh seamount basalts. These analyses are discussed under "Alteration of the Basalts" (p. 149) and "Chemical Analyses of Basaltic Glass" (p. 146). Comparative Discussion of the Chemical Results The similarity among the chemical compositions of the fresh seamount chain basalts is indicated by the small standard deviations of the means for each oxide of the ten analyses (Table VI, Col. 3). It is also demonstrated by the close grouping of the analyses on the Na20 + K20 - FeO (total iron) - MgO ternary plot (Fig. 40a, p. 139). This cluster coincides with the main grouping of basalts from Juan de Fuca and Sovanco Ridges (Fig. 40b, p. 139) and also with the compositional field of ocean ridge basalts*. However, the cluster is far-removed from the position of the average ocean alkali (from Engel ejt a l , 1965). *This field includes seventy-eight analyses of ocean ridge basalts up to mid-1969; after Miyashiro et a l , 1970. 132 TABLE VI. Mean partial chemical analyses and standard deviations of non-porphyritic (Col. 1) and porphyritic (recalcu-lated, Col. 2) basalts from Juan de Fuca and Sovanco Ridges, and of basalts from the seamount chains (Col. 3) 1 2 3 SiC-2 48 .9 ± 1. 0% 48. 4 ± 0. 4% 50. 0 ± 0. 8% A1203 15 .4 0. 5 16. 6 0. 6 14. 8 0. 3 Ti02 1 .21 0. 06 0. 96 0. 09 1. 02 0. 16 FeOT 8 .8 0. 3 8. 7 0. 5 8. 6 0. 5 MgO 7 .4 0. 2 7. 5 0. 5 7. 4 0. 4 CaO 11 .2 0. 1 11. 7 0. 4 12. 0 0. 4 Na20 2 .4 0. 2 2. 0 0. 1 2. 3 0. 3 K20 0 .07 0. 03 0. 10 0. 03 0. 08 0. 03 MnO 0 .16 0. 005 0. 15 0. 01 0. 17 0. 008 Total 95 .54 96. 11 96. 37 FeOj = Total Fe as FeO Col. 1. Non-porphyritic Type A basalt from Juan de Fuca and Sovanco Ridges. Mean of 4 analyses. Col. 2. Porphyritic Type A basalt from Juan de Fuca and Sovanco Ridges. Mean of 7 analyses. Six analyses were, recal-culated to remove large modal concentrations of plagioclase or olivine xenocrysts. Col. 3. Seamount chain basalt. Mean of 10 analyses from six dredge stations on Heck and Heckle Seamount Chains. 133 The most obvious difference between the fresh seamount chain basalts and the basalts from Juan de Fuca and Sovanco Ridges in the study-area is the chemical diversity of the latter basalts compared to the former, and this is demonstrated on a l l the chemical plots. The mean of four analyses of non-porphyritic basalt from Juan de Fuca and Sovanco Ridges is given in Table VI. The standard deviations are small. Also shown is the mean of seven analyses of porphyritic basalt from the same features, six of which were recalculated to remove modal concentra-tions of plagioclase or olivine xenocrysts. Considering the uncertainties of this recalculation as previously discussed (p. 126) the similarity between the mean analyses of the non-porphritic and recalculated porphyritic basalts suggests that they represent one type (referred to in this thesis as Type A), the porphyritic members of which have been modified by crystal accumulation. The porphyritic plagioclase basalts are not considered to be high-alumina basalts as described by Miyashiro e_t al_ (1969) from the Mid-Atlantic Ridge because the less porphyritic specimen of the same type (specimen 1-2) does not show exceptionally high alumina, and because the plagioclase crystals have characteristics of xenocrysts. The differences between the other basalts from Juan de Fuca and Sovanco Ridges and this predominant type are small. Specimens 2-2 and 2-3 with their higher potash and titania contents and their lower solidification indices are probably the product of minor differentiation of the predominant type. However, the origin of the high soda, low potash and titania basalt of Station 4 is not understood. The two analyses from Station 6 show minor differences from the predominant ridge basalt and from one another, although specimen 6-1 with its high potash and titania contents might be related to specimen 6-2 by differentiation and/or weathering. 134 None of these variants of Juan de Fuca and Sovanco Ridge basalts are identical to seamount chain basalts in major element composition, although the differences are small. Three oxides show minor systematic differences between the ridge basalts and the seamount chain basalts. Compared to the ridge basalts, the lime contents of the seamount chain basalts tend to be higher and the alumina contents lower (Fig. 41b, p. 141). Also the silica contents tend to be higher in the seamount chain basalts (Fig. 37). Despite these small variations, a l l the seamount chain and ridge basalts from the study area are ocean ridge basalts on the basis of chemical composition, and their variations f a l l within the range recorded for such basalts from the world's oceans. None of the fresh basalts from the study area are alkali, with the possible exception of specimen 4-1, because of its anomalous soda content. However, its other oxides have closer affinities with ocean ridge basalts. The basalts from the study area tend to be lower in alkalies (Fig. 37) and titania than the average ocean ridge basalt, but basalts of similar composition have been reported (for example, East Pacific Rise and Gorda Ridge--Kay et_ a l , 1970; Mid-Atlantic Ridge--Nicholls et a l , 1964). The uranium contents of the basalts from the study area are correspondingly low, generally less than 0.01 ppm, compared to about 0.2 ppm for the Mid-Atlantic Ridge near 45°N (see Table X, p. 166) (Aumento, 1969). Petrogenesis of the Basalts As more and more analyses of ocean ridge basalts are completed, evidence suggests that systematic differences may exist, both among basalts from one region and among basalts from different areas. 135 A JUAN DE FUCA RIDGE A SOVANCO RIDGE • HECK SEAMOUNT CHAIN • HECKLE SEAMOUNT CHAIN $ AVERAGE OCEAN THOLEIITE O AVERAGE OCEAN ALKALI (From Engel et a l , 1965) 6 r -o CM + 3 o CM (0 Hawaiian ^ Alkali / / / Hawaiian Tholelitlc S. A / A A A x x x X X 1*3 h5 ^6 k7 U9 50 °/o Si02 51 52 53 5* 55 FIGURE 37. Alkali/silica diagram of the basalts from the study area. Also shown are the average ocean alkali and average ocean tholeiite from Engel et al_ (1965), and the Hawaiian alkali-tholeiitic division line from Macdonald and Katsura (1964). 136 Aumento (1968) found a fractionation trend from olivine tholeiite to high-alumina tholeiite to quartz tholeiite among basalts from the Mid-Atlantic Ridge near 45°N. Miyashiro et_ al_ (1969) suggested that "abyssal tholeiites" (ocean ridge basalts and gabbros) from the Mid-Atlantic Ridge near 24° and 30°N had undergone crystallization differentiation, resulting in regular compositional variations even within single dredge hauls. Miyashiro et al_ (1970) found a well-developed iron-enrichment differentiation trend among tholeiitic gabbros from the Mid-Atlantic Ridge at 24°N (Fig. 40c). Kay et_ al_ (1970) observed an iron-enrichment trend among basalts from Juan de Fuca Ridge which they thought to be due to plagioclase and olivine fractiona-tion. Hence, although remarkable for their similarity, ocean ridge magmas have undergone differentiation in places. However, on the basis of the existing chemical analyses, the basalts from the study area do not show any well-developed differentiation trend (Fig. 40a, b), although the recovery of xenocrystic basalts implies that at least some fractionation has occurred. The range of solidification indices is small for both seamount chain and ridge basalts, and meaningful trends cannot be drawn (Fig. 38 and 39). Only three specimens from Juan de Fuca and Sovanco Ridges have somewhat lower solidification indices and/or higher Ti02 and K20 contents, suggesting that they may be the products of minor differentiation. When the study area data are considered in combination with chemical analyses from Dellwood Seamount area 200 km to the north, a trend which coincides with the Hawaiian alkali basalt trend (MacDonald and Katsura, 1964) is observed (Fig. 40b). The Dellwood Seamount area is a possible locus of sea floor spreading (Bertrand, 1972). The two Dellwood analyses which lie within the ocean ridge basalt field and closest to the McO corner of the AFM 137 • • • • m o a a . .a n&-m -51 -50 CM O -^9 co 16 15 0 ^ CM , f-t 13 o -112 «" 11 10 0) _ 8^ -» 7 8 1 o 60 7s J 2-J is Jo . i E 3V »aJ3, "75 33 50" SOLIDIFICATION INDEX In -rfc KtrO x 100 MgO + FeO + Na20 + KgO -.0.1 Jo.o CvJ FIGURE 38. Oxide concentrations (Wt. %) vs. S. I. for seamount chain basalts (from Table IV). Dashed lines represent mean concentrations of each oxide (from Table VI, Col. 3). 138 • o o o 52 51 50 49 48 J 47 C N o • H CO to O C N o E—' O CD O I? E-A A JO D « o A A A * A O *- . o . o 1 34 36 38 40 42 SOLIDIFICATION INDEX 44 46 48 50 52 MgO x 100  MgO + Fe0T + Na20 + K20 -] 3 2 1 2 1 0. 0. 0. 0. 54 o C N C N o o o C N 0.0 FIGURE 39. Oxide concentrations (Wt. %) vs. S. I. for ridge basalts (from Table III). Open symbols represent Sovanco Ridge basalts; closed symbols, Juan de Fuca. Dashed lines represent concentrations of each oxide in mean non-porphyritic Type A ridge basalt (from Table VI, Col. 1). 139 FEO a. • SEAMOUNT BASALTS (10 analyses) O AVERAGE OCEAN ALKALI BASALT (af t e r Engel et a l , 196S) 0 FIELD OF OCEAN RIDGE BASALTS (78 analyses to mid-1969; from Miyashiro et al_, 1970) FEO NA2O+/ K 20 ' ASOVANCO RIDGE BASALTS (7 analyses) •JUAN DE FUCA RIDGE BASALTS (10 analyses) ODELLWOOD SEAMOUNT AREA BASALTS (13 analyses from Bertrand, 197 FEO j/ SKAERGAARD TREND ^'HAWAIIAN THOLEIITE TREND / HAWAIIAN ALKALI TREND MGO GORDA RIDGE BASALTS (7 analyses) (from Kay et a l , 1970) JUAN DE FUCA RIDGE BASALTS (7 analyses) (from Kay ejt al_, 1970) SOVANCO RIDGE BASALTS (7 analyses) (study area) JUAN DE FUCA RIDGE BASALTS (10 analyses) (study area) FIELD OF OCEAN RIDGE BASALTS (78 analyses to mid-1969; from Miyashiro et a l , 1970) ( 'IRON-ENRICHMENT TREND (MID-ATLANTIC RIDGE) ^- (from Miyashiro et a l , 1970) MGO FIGURE 40. (Na20 + K20) - FeO (total iron) - MgO diagrams for basalts from the study area (a.Heck and Heckle Seamount Chains; b. and c, Sovanco and Juan de Fuca Ridge basalts) and other basalts and gabbro. 140 triangle are from the probable spreading centre, whereas the other analyses represent either seamount basalts or weathered basalts. Weathering can produce trends similar to differentiation trends (see p.154). Obviously the basalts from the Dellwood Seamounts, which are generally alkali or transitional to alkali (Bertrand, 1972), contrast with those from the seamount chains in the study area, and are perhaps related to the spreading centre basalts in a way similar to that described by Aumento (1967) for the alkali basalts from seamounts associated with Mid-Atlantic Ridge. A more meaningful comparison is between analyses from the study area and analyses from Juan de Fuca and Gorda Ridges to the south (analyses from Kay et_ a l , 1970). The combined data show an iron-enrichment trend similar to that observed by Miyashiro et_ al_ (1970) and to the Skaergaard trend (Fig. 40c). As demonstrated by Kay et al (1970) these trends could be produced by fractional crystallization of forsteritic olivine and calcic plagioclase (Fig. 41a, b). Such fractionation also explains the rare-earth element abundances and nickel and strontium trends in these basalts, which were determined by Kay et_ al_ for their specimens. They concluded from their chemical data that shallow fractional crystallization yielding iron-enriched differentiates has occurred in some basalts from Juan de Fuca Ridge. They also concluded that the parent liquids had alumina contents between 16 and 17.5%, and contained hypersthene and olivine in their norms, low amounts of elements with large cations, and unfractionated rare-earth element patterns. Basalts from the study area are chemically similar to those from Gorda Ridge except the alumina contents tend to be lower in the study area basalts (Fig. 41b). In addition, a higher proportion of the study area basalts have quartz in their norms (11 out of 27 compared to 1 out of 6, using the ratio Fe3+/Fe^+ = 0.1 to calculate the norms). As will be discussed below, this 141 FIGURE 41. a. FeO against'.MgO and b. CaO against A I 2 O 3 (Wt. %) for basalts from the study area. Juan de Fuca (44°N), Gorda (41°N), and Mid-Atlantic (45°N and 30°N). Ridge fields are from Kay et al (1970). Lines joining open circles show trends resulting from removal of plagioclase and olivine from a high alumina basalt (from Kay et a_l, 1970). Legend applies to both a. and b. 142 might be due either to a slightly shallower depth of origin for the study area basalts, or greater fractional crystallization of the study area basalts. Bass (1971) compiled and compared the major element compositions of basalts from different ridge crests, and suggested that systematic differences exist in their normative compositions. To calculate the norms, he used a constant ratio of Fe^+ to Fe^+ of 0.1, which is virtually the lowest ratio determined in basalts, and probably the best approximation to the mantle ratio. His results were plotted using anorthite instead of the more conventional diopside because diopside is a secondary normative quantity, derived after the calcium for anorthite has been deducted, whereas anorthite is a direct measure of aluminum. Aluminum may be a critical parameter of the chemical variation of basalt. The fields described by Bass for ridge basalts are shown in Figure 42b. Superimposed is the field from Figure 42a for non-porphyritic basalts from the study area, the norms of which were also ~ 2+ calculated using a constant Fe-5"1" to Fe ratio of 0.1. Bass concluded on the basis of his compiled data that variations in the normative compositions of ocean ridge basalts reflect spreading rate. The Mid-Atlantic Ridge is considered to be a slow-spreading ridge, Gorda, moderately slow, and Juan de Fuca Ridge and the East Pacific Rise relatively fast. With decrease in spreading rate, the fields of basalts shift progres-sively toward the An-01 side from the An-Hy side. This writer considers that Bass' conclusions are not valid because the Mid-Atlantic Ridge field completely encloses the East Pacific Rise field. However, additional analyses may result in a statistically better division because the extension of the East Pacific Rise field towards the An-01 join is the result of only one analysis; two of the others lie within Bass' Juan de Fuca Ridge field and the third just outside i t . 143 + SCV.RNCG RIDGE 9flS3LT * JUfiN DE FllCfl RIDGE BflSPLT • SEAMOUNT SfiSflLT / FIELD OF NON - PORPH1RITIC BASALTS from the study area b. RIDGE BASALTS MID - ATLANTIC RIDGE ( > 50 analyse EAST PACIFIC HISS (1* analyses) GORDA RIDGE (6 analyses) JUAN DE FUCA RIDGE (9 analyses) THIS STUDY (19 analyses fromi Juan de Fuca Ridge Sovanco Ridge Seamount Chains) SEAMOUNT BASALTS JUAN DE FUCA RIDGE (2 analyses) EAST PACIFIC RISE C* analyses) HID - ATLANTIC RIDGE (19 analyses) FIGURE 42. a. Normative mineralogy of basalts from the study area, plotted on a nepheline-olivine-anorthite-hypersthene-quartz diagram, b. Fields of various submarine basalts on normative mineralogy diagram as in a. Field of non-porphyritic basalts from the study area is taken from a. Other fields and seamount basalts are from Bass (1971). 144 The basalts from Juan de Fuca Ridge in the study area coincide more closely with the Gorda Ridge field than the Juan de Fuca, further indication that the trend postulated by Bass is not valid. However, the situation is complex because the position of Bass' Juan de Fuca field is probably largely controlled by fractionation, as his data are from Kay ejt al_ (1970). Also, the spreading rate near the northern end of Juan de Fuca Ridge is somewhat equivocal (see Chapter V). Unless confirmed by more chemical analyses from the Pacific Ocean, Bass' conclusions are statistically suspect. However, on the basis of available data, this author suggests that there may be a significant difference between the Mid-Atlantic Ridge basalts (whose field is well-defined by about 50 analyses from many areas and by many different analysts) and the basalts of Juan de Fuca and Gorda Ridges, in that the field of the former basalts does not enter the quartz-normative triangle on Figure 42b. The theories for the origin of basalt magmas were discussed in Chapter II of this thesis. It was noted that experimental work (for example, Green and Ringwood, 1967) has indicated that the compositions of the principal basalt magmas may be determined by two different pressure-dependent variables: 1. Depth in the mantle at which magma segregation occurs 2. Depth at which fractionation (if any) occurs. Quartz tholeiite may be derived either by fractional melting and segregation from pyrolite in the depth interval 0-15 km, or by crystallization and segregation of olivine from an olivine tholeiite magma in the same depth interval. Similarly, the typical olivine-normative ocean ridge basalt may be derived as a direct product of magma segregation from pyrolite in the interval 15-35 km, or by fractionation of an olivine tholeiite in the same depth interval. Because of the compositional similarity of ocean ridge basalts, 145 many workers have concluded that they represent the product of mantle segregation followed by rapid rise to the surface without appreciable fractionation. Because quartz tholeiites are quantitatively minor on the Mid-Atlantic Ridge, Aumento (1967) concluded that they are the product of frac-tionation. However, on the Pacific Ocean ridges basalts commonly tend towards the quartz tholeiite field. Hence, i t is reasonable to speculate that they are not solely the products of shallow differentiation, but are derived from shallower depths than the Mid-Atlantic Ridge basalts, and that the faster-spreading ridges or at least Pacific Ocean ridges, do tap magma from shallower depths (close to 15 km) than the slow-spreading M. A. R. The elongation of the fields in Figure 42b probably reflects dif-ferentiation, but, as suggested by Bass, may also reflect the tapping of magmas from different depths down to the maximum for a particular ridge. Also shown in Figure 42b is the composition field for M. A. R. seamounts and the compositions of six seamounts from Juan de Fuca Ridge and the East Pacific Rise. Obviously the basalts from the Heck and Heckle Seamount Chains have an origin different from these other seamount basalts, and are closely related to the ridge basalts. Their normative compositions span the olivine-quartz boundary. Since there is no evidence for fractiona-tion of the seamount basalts in the form of xenocrystic specimens or variation among the analyses, their compositions are further evidence that basalts in the study area are primarily the products of magma segregation at depths of about 15 km, rather than of extensive fractionation. In conclusion, the basalts from the study area, including those from the seamount chains, are compositionally ocean ridge basalts which are similar to other basalts from Juan de Fuca and Gorda Ridges or which can be related to them by fractional crystallization processes. Within the study area only 146 minor differentiation has occurred at shallow depths. The magmas may originate at shallow depths (about 15 km) by partial melting of and segregation from the mantle. This is in accord with the high geothermal gradient and upraised mantle under the ridge (Dehlinger et_ a l , 1970). Chemical Analyses of Basaltic Glass Five samples of basaltic glass from the study area were analyzed. In all cases fresh-looking, lustrous black glass was sampled, and the outer yellow-brown palagonitized glass was avoided. The results of the analyses are shown in Table VII. Four of the samples represent the glass margins of samples whose more crystalline interiors were also analyzed, and these latter analyses are presented for comparison. For the two specimens from Juan de Fuca Ridge (3-1 and 4-1) the analyses of the glass margins do not differ significantly from the analyses of the interiors. However, for the two specimens from Heck Seamount (8-1 and 10-1), the glass margins show a large increase in alumina and decreases in a l l the other oxides relative to the interiors. The fragment of glass from Station 9 (9-1) contains an abnormally large alumina content, and depleted concentra-tions of a l l the other oxides but most notably s i l i c a , relative to the average seamount chain basalt. This analysis is the mean of two duplicate analyses, reported in Appendix B, Part 2. The comparative analyses of interiors and glass margins of pillows and vitric breccias associated with pillow basalts have been discussed TABLE VII. Partial chemical analyses of fresh basalts and their glass selvages (Col. 1-8). Specimen 9-1 (Col. 9) represents a small piece of glass with which no crystalline basalt was associated. Columns 1-4: Juan de Fuca Ridge; columns 5-9: Heck Seamount. 1 2 3 4 5 6 7 8 9 Specimen Number 3-1 3-1G1 4-1 4-1G1 8-1 8-1G1 10-1 10-1G1 9-1G1 Si02 47. 7 48.1 49.1 48.1 50.7 46.0 50.9 49.7 42.3 A1203 15. 6 15.3 15.7 16.7 15.0 23.8 14.8 18.4 27.6 Ti02 0. 75 0.79 0.88 0.88 0.87 0.80 0.98 0.85 0.82 FeOT 8. 1 8.2 8.3 8.6 8.8 8.1 9.6 9.2 8.4 MgO 11. 8 12.4 7.6 7.9 8.2 7.5 7.8 7.3 6.6 CaO 10. 7 10.6 12.0 12.1 11.8 10.9 11.8 11.0 10.1 Na20 2. 0 2.4 3.6 5.0 2.6 2.1 2.4 2.0 1.7 K20 0. 04 n.d. 0.04 0.05 0.04 0.03 0.09 0.03 0.02 MnO 0. 15 0.16 . 0.16 0.17 0.17 0.16 0.18 0.18 0.16 Total 96. 8.4 97.95 97.38 99.50 98.18 99.39 98.55 98.66 97.70 Fe0T = Total Fe as FeO 148 occasionally in the literature (for example, Bailey et_ a l , 1964; Vallance, 1965; Aumento and Loncarevic, 1969; Miyashiro et a l , 1969; Muffler et a l , 1969), and a l l results do not show the same trends. Specimens 3-1 and 4-1 from Juan de Fuca Ridge appear to be typical of young, fresh pillow basalts from the ocean floors, in which interiors and glass margins have generally been reported to be rather similar in at least major element composition. Exceptionally, in submarine basalts from abyssal h i l l s in the South Pacific Ocean, Paster (1971) found a chemical trend from glass and variolitic layers to the crystalline cores of pillows involving decreases in total iron and magnesia and increases in the other major oxides. In at least one case, the changes could be associated with deuteric alteration involving the formation of secondary chlorite in the interior. Hence, Paster considered the glass and variolitic layers to reflect the original chemistry more than the interior. However, this cannot be the case for the Heck Seamount specimens whose interiors have the typical seamount chain basalt composition. Interestingly, of the five specimens, only 3-1 and 4-1 are typical pillow fragments. Specimens 8-1 and 10-1 are irregular in shape, probably the products of more violent eruptions, and specimen 9-1 is only a small slab of glass. Comparative analyses of glasses and interiors of equivalent sea floor samples were not found in the literature. Most analyses which show marked differences in chemical composition between cores and glass margins are from pillow basalts of older geologic ages now exposed on the land, and it has been suggested that such differences were produced by weathering and metamorphism (Miyashiro et a l , 1969). However, although Bailey e_t al_ (1964) admitted that how and when the changes were brought about were not known, they suspected that significant changes were brought about by i n i t i a l reaction of the molten magma with sea water. 149 Unfortunately, except for the apparent loss in s i l i c a , the trends documented in these older pillows are unlike those of the Heck Seamount samples, par-ticularly in that the alumina contents tend to remain roughly constant. Furthermore, specimens 8-1, 10-1, and 9-1 have not been metamorphosed or extensively weathered, and the interiors of specimens 8-1 and 10-1 are chemically typical of fresh seamount chain basalts. The alteration of basaltic glass to palagonite involves loss of s i l i c a , magnesia, lime, and soda, and gain in potash, titania, and iron, as well as hydration (Moore, 1966; Matthews, 1971; Paster, 1971). However, alumina remains roughly constant. Consequently, palagonitization does not seem likely to cause the high alumina trends of the Heck Seamount glasses. Because similar compositional trends are shown by four obviously altered specimens from Heck Seamount Chain (see p. 154), the changes are thought to be the results of deuteric alteration, i n i t i a l sea water alteration, early stages of submarine weathering, or combinations-of these. The exact processes involved are not known, and a detailed investigation of them is beyond the scope of this study. In conclusion, these results indicate that large differences can exist between glassy margins and more crystalline interiors of submarine basalt specimens. This difference may be more likely in the case of non-pillowed basalts. The often-made assumption that glass margins are more likely to represent the original magma composition is not always valid. Alteration of the Basalts The basalts from the study area are generally quite fresh and alteration products are quantitatively not very important. However, because alteration and submarine weathering can produce important chemical changes 150 in basalt (for example, Matthews, 1971) it is of interest to describe briefly the products and chemical effects observed in some of the rocks from the study area. Broadly speaking, four types of alteration or weathering were observed to have affected the basalts from the study area: a. Palagonitization b. Replacement by serpentine and/or chlorite and associated materials c. Formation of chlorophaeite and associated materials in the rims of basalt fragments d. Irregular alteration throughout small basalt fragments Descriptions of these materials are presented in Table VIII. The products of alteration and weathering are extremely variable in appearance and difficult to study optically because of their generally cryptocrystalline natures. Consequently the classification given in Table VIII may be over-simplified. The chemical effects of the formation of altered rims in which vesicles and interstices contain chlorophaeite and iron oxides and of irregular alteration throughout basalt fragments were investigated as part of this study.' Unfortunately, two of the most sensitive indicators of the degree of alteration and weathering--the ferric/ferrous iron ratio and the total water content-were not determined. Altered Rims Samples of both the interiors (light to medium grey) and dark grey rims of six specimens were analyzed to determine some of the chemical changes brought about by this type of alteration. The results are shown in Table IX. The general results are summarized below, from interior to rim: TABLE VIII. Summary of major types of alteration observed in basalts from the study area TYPE FORM AND APPEARANCE ACCOMPANYING CHEMICAL CHANGES Palagonitization Yellow-brown crust on glass, or yellowish matrix in breccias. In thin section, generally yellow; may be either isotropic gel-palagonite or bire-fringent fibro-palagonite (Peacock, 1926; Peacock and Fuller, 1928). Increase in: water, K20, Ti02, and total Fe. Decrease in: Na20, CaO, MnO, MgO(?) and Si02(?) (Moore, 1966; Matthews., 1971; Paster, 1971). Replacement by serpentine and/or chlorite and assoc. materials In some cases there is a brown tint to the hand specimen. In thin section, green or brown cryptocrystalline or fibrous alteration product occurs in interstices, or pseudomorphic after olivine. Carbonates and iron oxides are usually associated. Occurs throughout specimen (not related to margins). Not documented. Formation of chlorophaeite and assoc. materials Specimens have dark rims, sometimes yellow- or brown-tinted. In thin section, bright green (in some specimens yellow or brown) material f i l l s or lines vesicles, occurs in irregular patches in the groundmass, or f i l l s veinlets in the dark rim only. Usually cryptocrystalline and weakly birefringent; in places fibrous and birefringent. Iron oxides are commonly associated. Increase in: total Fe, K2O, Na20(?). Decrease in: MgO Generally l i t t l e change in Si02, AI2O3, Ti02, Cao, or MnO, but in some cases they show small losses. Irregular alteration Affects cryptocrystalline basalt. Vari-coloured discolouration on hand specimen, in some cases related to vesicles or cracks. In thin section, green or yellow "chlorophaeite" and iron oxides are observed in some bands, but banding is also caused by varying turbidity of the groundmass. Increase in: AI2O3. Decrease in: All other major oxides. Similar chemical changes may also affect glass margins of non-pillowed basalts. TABLE IX. Partial chemical analyses of fresh interiors (Col. 1, 3, 5, 7, 9, 11) and altered rims (Col. 2, 4, 6, 8, 10, 12) of basalts from Juan de Fuca Ridge (Col. 1-4), Heck Seamount Chain (Col. 5, 6) and Heckle Seamount Chain (Col. 7-12). Columns 13-16 are analyses of altered basalts from Heck Seamount Chain 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Specimen Number 1-1 1-1AR 2-2 2-2AR 7-1 7-1AR 12-1 12-1AR 12-2 12-2AR 12-3 12-3AR 11-2 11-3 7-3 7-2 Si02 47.4 47.9 52.1 48.8 50.8 48.7 50.5 50.1 49.0 47.7 49.5 48.4 47.6 47.5 44.6 52.0 A1203 19.0 20.0 14.4 12.8 15.1 14.3 14.6 13.8 15.7 14.7 14.7 14.6 20.6 21.8 23.9 16.9 Ti02 0.83 0.65 2.08 1.89 1.00 0.97 1.22 1.13 1.33 1.25 1.22 1.21 0.74 0.75 0.66 0.72 FeOr 7.3 6.8 9.3 11.9 8.4 9.9 8.3 10.2 8.0 10.4 8.2 9.8 6.4 7.8 7.3 7.0 MgO 6.4 5.1 • 7.5 6.3 7.6 7.0 7.0 7.1 7.0 6.3 7.2 7.4 6.8 6.6 7.6 5.8 CaO 12.8 12.8 10.6 10.2 12.5 11.3 11.6 11.1 11.6 11.0 11.5 11.3 11.9 11.0 11.1 11.8 Na20 1.9 2.5 3.2 2.8 2.1 2.1 2.1 2.1 2.6 2.5 2.5 2.7 2.1 2.0 1.7 2.0 K20 0.08 0.17 0.19 0.22 0.05 0.46 0.08 0.34 0.11 0.35 0.09 0.06 0.06 0.13 0.05 0.16 MnO 0.13 0.10 0.21 0.19 0.16 0.19 0.18 0.18 0.18 0.18 0.18 0.20 0.12 0.14 0.13 0.14 Total 95.84 96.02 99.58 95.10 97.71 94.92 95.58 96.05 95.52 94.38 95.59 95.67 96.32 97.72 97.04 96.52 FeOj. " Total Fe as FeO 153 S 1 O 2 Small loss, or l i t t l e change AI2O3 Small loss, or l i t t l e change Ti02 Small loss, or l i t t l e change FeO Large gain, or l i t t l e change MgO Loss, or l i t t l e change CaO Little change, or a small loss Na20 Little change, or a gain K20 Large gain MnO Little change Although data in the literature is sparse, these results appear to be consistent with results obtained from other areas on the submarine weathering of basalt. Matthews (1971) found that the altered basalts from Swallow Bank in the eastern Atlantic Ocean show no change in MnO, possible loss of s i l i c a , marked loss of magnesia and lime, and marked gain in potash, as well as oxidation of iron and increase in water content, compared to fresh basalts. As noted by Matthews, the constancy of manganese does not support an origin for the manganese of manganese encrustations and nodules in the submarine weathering of basalt. Miyashiro et_ al_ (1969) found a decrease in silica and an increase in total iron, as well as an increase in the oxidation of iron and the water content within one specimen in successive layers from the interior to the exterior. But the variations in magnesia, lime, soda, and potash are compli-cated and show maxima and minima in internal layers. Aumento et_ al_ (1971) in basalts from the Mid-Atlantic Ridge at 45°N found that lime is hardly changed with weathering, although there is a slight tendency for loss on severe weathering, whereas potash is markedly increased. 154 Because of the tendency for loss of magnesia, and gain of iron and potash, these changes could be mistaken for the effects of differentiation. However, consideration of oxides like silica and titania, as well as water content, oxidation state of iron, and the appearance of the basalt in hand specimen and thin section should make the distinction clear. Irregular Alteration Four altered basalts of this type were analyzed, and the results are shown in Table IX. Specimen 7-2 is a fragment of clinker, and less obviously altered than the other specimens. Compared to the typical fresh seamount basalts, these specimens show the same increase in alumina and decrease in the other oxides observed in the analyses of the seamount glasses previously described (Table VII). Similar trends have not been reported for submarine basalts. Because the glass rims of basalts whose crystalline interiors have average compositions show this same trend, some type of alteration or weathering process would seem to be reasonable, rather than a different type of parental magma. Palagonitization might be partly responsible in that i t can result in apparent loss in s i l i c a , iron, magnesia, and lime as already discussed. However, alumina apparently remains approximately constant during palagoniti-zation. Furthermore, palagonite does not seem to be a major component of either the altered basalts or the analyzed glass selvages of the Heck Seamount basalts. Although the altered basalts do contain chlorophaeite-like green material in interstices and vesicles, as well as iron oxides, the chemical trends are unlike those of the dark rim type of alteration previously described. The origin of this high alumina trend is an enigma. 155 Hyaloclastites and Aquagene Tuffs Breccias and tuffaceous material are commonly intimately associated with outcrops of pillowed volcanics of submarine origin (for example, Peacock, 1928; Silvestri, 1963; Carlisle, 1963). In six dredge hauls from the study area, similar breccias and tuffs were recovered, and their origins can be inferred by comparisons with the likely origins of their counterparts which are now readily observed in the geological record on the continents. Hyaloclastites As already mentioned, many basalt fragments from Stations 2 and 6 are partly or wholly encased in a "palagonite breccia" (Fig. 34b). Such fragments may have been part of pillow breccias, as described by Carlisle (1963). The proportions of breccia within the eruptive sequences described by Carlisle on Quadra Island, British Columbia, vary from less than 10% in the case of close-packed pillows to much higher in the isolated and broken-pillow breccias. Because of the angularity of most breccia-encased fragments from Stations 2 and 6, their most likely source is a broken-pillow breccia. Like those described by Carlisle, these breccias consist of angular basalt fragments set in a matrix of granules and shards of glass, plagioclase, pyroxene, and olivine crystals, ferruginous material, and various alteration products. The matrices of close-packed pillows and isolated-pillow breccias would be the same, but rounded, more complete pillow forms would be expected among the basalt fragments. The matrices of the close-packed pillows and of the pillow breccias are essentially equivalent to the "hyaloclastites" of Silvestri (1963) and the "palagonite tuffs" of many earlier writers, and hereafter are referred to as "hyaloclastites1.'. 156 Several fragments of hyaloclastites were dredged at Station 1. They consist of chips of glass and/or basalt fragments set in a yellowish matrix of palagonite, ferruginous material, and alteration products (Fig. 43a). In thin section, the vitric hyaloclastite is observed to consist of sidero-melane chips with palagonitized rims set in a matrix of yellow palagonite and ferruginous material (Fig. 43b). Anhedral plagioclase crystals and occasional pyroxene and olivine crystals are also floating in the matrix. The palagonite is predominantly birefringent fibro-palagonite. Other types of hyaloclastites were recovered. At Station 5, several large pieces of weathered breccia were obtained, consisting of basalt fragments about 1 inch across, and small chips of glass set in a matrix which was probably originally palagonite (and in part s t i l l i s ) , but which is generally altered to greyish-brown clays and oxides (specimen 5-7). At Station 11 a similar but less weathered breccia was recovered which consists of fragments of tachylite and basalt of varied size, shape, and texture in a palagonitic matrix with its typical yellowish hue. Anhedral plagioclase crystals and dark, ferruginous material are also in the matrix (specimen 11-4). Origin of hyaloclastites.--The unsorted pillow breccias and hyaloclastites are thought to form during the submarine extrusion of lava by the spalling off of glass from pillows, break-up of pillows, and also from flow of pillowing lavas during extrusion. Aquagene Tuffs Small pale brown pieces of very fine-grained tuffaceous deposits, containing flecks of black glass, were dredged from Stations 9 and 10 on Heck Seamount. In thin section, these deposits are seen to consist of shards of FIGURE 43. Hyaloclastites and tuffs from Juan de Fuca Ridge and Heck and Heckle Seamount Chains a. Vitric (on left) and basalt-fragment hyaloclastites from Middle Ridge. Specimen 1-6 and 1-9. b. Photomicrograph of vitric hyaloclastite, showing sideromelane and plagioclase fragments in a palagonitic and ferruginous matrix. Specimen 1-6. Plane light X45. c. Fragments of basalt and glass in a tuffaceous matrix. Specimen 12-4. d. Graded bedding in a laminated vitric tuff. Specimen 9-2. e. Laminated vitric tuff. Specimen 9-3. f. Photomicrograph of laminated vitric tuff, showing palagonitized sideromelane and tachylite fragments in a matrix of palagonite and ferruginous material. Specimen 9-3. Plane light X40. 157 FIGURE 43. H y a l o c l a s t i t e s a n d t u f f s f r o m J u a n d e F u c a R i d g e a n d H e c k a n d H e c k l e S e a m o u n t C h a i n s . 158 sideromelane, altered glass, a few fragments of plagioclase crystals, and a ferruginous matrix. A crumbly, tuffaceous fragment of breccia, consisting of scattered fragments of glass and basalt in a pale brown matrix (Fig. 43c) was recovered from Station 12 on the Heckle Seamount Chain. X-ray diffraction analysis indicated that the matrix is composed largely of amorphous material (probably ferruginous), as well as the manganese mineral todorokite, poorly crystalline ferruginous minerals, chlorite group minerals, and detrital material (quartz and plagioclase). This composition suggests that the deposit may be more closely related to the ferro-manganese deposits than to aquagene tuffs. Many pieces of laminated vitric tuff were recovered from Station 9 on Heck Seamount. These tuffs may form much of the sea floor at Station 9, as some have sponges growing on one surface, or patches of oxides. These tuffs consist of chips and shards of glass, as well as some small fine-grained basalt fragments, embedded in a fine-grained pale yellowish-brown matrix (Fig. 43d). They generally show graded bedding. One slab is darker in colour but more distinctly banded (Fig. 43c). A thin section of this slab shows that the glass chips are either dark tachylite or pale green sideromelane; both types are rimmed by yellow fibro-palagonite (Fig. 43f). The fibro-palagonite is birefringent under crossed nicols (Fig. 44a, b), whereas usually both tachylite and sideromelane are isotropic, although the tachylite may be slightly birefringent. The matrix between the glass fragments contains plagioclase, and more rarely olivine and clinopyroxene, crystals and small glass chips in a matrix of fibro-palagonite and dark ferruginous material. Origin of aquagene tuffs.--Similar laminated vitric tuffs occur within the "broken-pillow breccia" sequence on Quadra Island, British Columbia (Carlisle, 1963). Carlisle concluded that they formed during FIGURE 44. Tuffs and ferro-manganese deposits from Heck and Heckle Seamount Chains a. Photomicrograph of laminated vitric tuff, same as Figure 43f. Specimen 9-3. Plane light X35. b. Same as a., under crossed nicols to demonstrate isotropism of both sideromelane and tachylite, and birefringence of palagonite (fibro-palagonite). Specimen 9-3. Crossed nicols X30. c. Cut surface of ferro-manganese nodule. Specimen 11-5. d. Cut surface of ferro-manganese pavement, containing basalt fragment. Specimen 13-2. e. Orange ferruginous cap on loosely consolidated sedimentary deposit. Specimen 10-2. 159 FIGURE 44. Tuffs and ferro-manganese deposits from Heck and Heckle Seamount Chains. 160 formation of the broken-pillow breccia from a mixture of pillows and hyalo-clastites flowing turbulently; such a turbulence would set a cloud of glass shards and other debris into suspension. The layered tuff would be the rhythmic condensate of such a cloud. The fragments of fine-grained tuff are equivalent to the slowest-settling layer of the laminated tuff (Fig. 43d). They represent the sand, s i l t , and clay-sized debris produced by interaction of lava and sea water. Ferro-manganese Deposits Many rocks dredged from the study area have ferro-manganese oxide coatings of varying thicknesses (for example, Fig. 36b, c, e). The thickness of the oxide coating can be used to give a rough estimate of the age of the basalt (see p. 163). The coatings are usually finely banded, and vary from brown to black. X-ray diffractograms of these coatings show that they consist almost entirely of amorphous material. That iron in ferro-manganese nodules is almost a l l present as amorphous material has been noted by recent workers (for example, Cronan and Tooms, 1969), and the same seems to be true of ferro-manganese coatings. Apparently much of the manganese is also present as amorphous material, although small amounts of todorokite were present in some samples. Also present in minor andy varying amounts are clay minerals of the chlorite and montmorillonite groups and detrital quartz and feldspar. Elemental analysis of a similar coating from the Southeast Dellwood Knolls 200 km to the north contained 22.1% Mn, 6.13% Fe, 8.1% S i , 2.2% a l , 1.8% Mg, 1.4% Na, L.2% Ca, 1.1% K, and 0.15% t i (Bertrand, 1972). This illustrates the large amount of manganese concentrated in these deposits A ferro-manganese nodule from Station 11 on the western peak of the Heck Seamount Chain consists of a core of altered, fine-grained to glassy, 161 brecciated basalt surrounded by discontinous, multi-coloured, concentric layering (Fig. 44c). For comparison, X-ray diffraction analyses were done on samples from three zones identifiable in Figure 44c. The inner orange-brown zone contains largely amorphous material (ferruginous and manganiferous) with minor todorokite and goethite, chlorite and montmorillonite group minerals, quartz, and plagioclase. The very dark coloured zone outside the orange-brown zone contains much more todorokite, as well as amorphous material, montmorillonite group minerals, quartz, and plagioclase. The crumbly outer zone of the nodule also contains considerable todorokite, with amorphous material, plagioclase, and quartz. Hence, considerable environmen-tal variation may have occurred during the deposition of the nodule. The quartz and plagioclase are detrital in origin. A fragment of ferro-manganese pavement very similar to some of those described from San Pablo Seamount in the North Atlantic Ocean (Aumento et_ a l , 1968) was recovered from Station 13 near the western end of the Heckle Seamount Chain. It consists of roughly botryoidal and irregular banding of black and orange-brown layers, with some embedded glass and basalt fragments (Fig. 44d). It lacks the depositional nuclei of the ferro-manganese nodules. X-ray diffraction analysis indicates that the pavement consists mostly of amorphous material, with minor goethite, todorokite, plagioclase, and quartz. Underwater photographs on San Pablo Seamount (Aumento et_ a l , 1968) show similar deposits completely paving the basalt of the seamount. The presence of such a pavement on the western regions of the seamount chains in the study area would explain why dredging in those regions was so unsuccessful. From Station 10, thinly laminated orange ferruginous deposits were recovered both as fragments and as cappings on loosely consolidated green sedimentary deposits (Fig. 44e). X-ray diffraction with Cu K<* radiation 162 shows the orange material to be amorphous, whereas the green deposit contains quartz and montmorillonite and chlorite group minerals. Ferro-manganese deposits form by chemical deposition at the sediment-water interface, or at the basalt-water interface. The source of the iron and manganese in these deposits has been the subject of debate, both the continents and the alteration of submarine basalt having been suggested. In addition, i t has been suggested that hydrothermal sources may be more or less important (Cronan and Tooms, 1969). In view of this controversy, detailed mineralogical and chemical studies of the ferro-manganese deposits from the study area would be valuable; however, such studies are beyond the scope of the present investigation. Ages of the Basalts The ages of the ocean floor in the study area have been predicted from the magnetic anomalies combined with the geomagnetic reversal time scale (Fig. 4). On the basis of this geochronologic tool, the oldest basalts that have been collected would be those from Station 13, near the western end of the Heckle Seamount Chain, which can have a maximum age of about 3.5 my. The youngest specimens (less than about 0.7 my) would be those from Stations 3 (West Ridge) and 7 (the eastern peak of the Heck Seamount Chain), and possibly those from Stations 6 (Sovanco Ridge), 4 (Endeavour Seamount), and 1 and 2 (Middle Ridge). Specimens from the other dredge hauls would be expected to be of intermediate age. In an attempt to determine the accurace of these predictions, two methods of directly estimating the ages of the dredged specimens were used: Ferro-Manganese Geochronology Fission Track Age Dating 163 The ferro-manganese geochronology reveals excellent qualitative agreement with the relative ages predicted by the geomagnetic time scale. Unfortunately, the basalts proved generally unsuitable for fission track age dating because of their unexpectedly low uranium contents. However, the two ages obtained can be interpreted to be compatible with the geomagnetic time scale. Ferro-manganese Geochronology As described in the section concerning ferro-manganese deposits, many of the specimens collected are covered by varying thicknesses of ferro-manganese coating. This coating is accreted by unknown mechanisms directly from sea water onto any exposed surface which is kept free of sediment. A plot of maximum coating thickness against the magnetic anomaly age for each dredge haul, excepting Station 6, in the study area (projected onto a line perpendicular to the anomalies through Station 4) shows a definite correlation between increasing thickness and increasing predicted age (Fig. 45). Notably, the only ferro-manganese nodule collected and the slab of ferro-manganese pavement were recovered in the two dredge hauls (Stations 11 and 13) furthest from West Ridge. Furthermore, the freshest basalts collected from the study area are those from West Ridge, Endeavour Seamount, and the eastern peak of the Heck Seamount Chain (see Petrography of the  Basalts). Admittedly these data are qualitative. Rates of ferro-manganese oxide coating accumulation on basalts have not been measured in the North Pacific. The rates of ferro-manganese accretion in nodules have been measured, and the minimum rate is 2.5 mm/10 years, and the maximum rate 40 mm/10^  years (Barnes and Dymond, 1967; Bender et a l , 1970). The majority of measurements lie near the minimum rate. The 164 I I I I 3.4 my 2.4 my 2.0 my H J < U CO A \ LU CO c n H i - H >-MA c • H o > AN u <L> u • P H-l CH E - ct) LU v / Z C J MA 2.0 my FERRO-MANGANESE COATINGS MAXIMUM THICKNESS (mm) 5 10 j ESTIMATED AGE (my) Max. Min. T 0.25 2.4 0.15 2.0 0.125 0.8 0.05 0.4 0.025 I FIGURE 45. Maximum thicknesses of ferro-manganese coatings on rocks from each dredge station plotted against the posi-tion of each station projected onto a line through Station 4 perpendicular to the magnetic anomalies. Magnetic anoma-lies shown are those along the line of projection. Also shown are the ages of the sea floor predicted by the magnetic anomaly time scale, and the estimated ages of coated specimens. The maximum is based on an accumulation rate of 40 mm/106 yrs., the minimum on a rate of 2.5 mm/106 yrs. 165 nodule recovered from Station 11 in the study area has a maximum oxide thickness of 15 mm, over twice the maximum coating thickness on associated basalt fragments. Hence, oxide coatings may accumulate more slowly than nodules. Aumento (1969) found two rates of oxide accretion on Mid-Atlantic Ridge basalts, 1.64 ± 0.2 mm/106 years and 4.1 ± 1.2 mm/106 years. To roughly estimate the ages of the basalts in the study area, the maximum and minimum rates of Pacific nodule accumulation were used, bearing in mind that even the minimum rate of nodule accumulation may exceed the rate of coating accumulation, and that the minimum rate (2.5 mm/106 yrs) is more likely than the maximum (40 mm/106 yrs). These rates rely on the assumptions that accumulation rates are constant and that at no time were the specimens buried. Despite the uncertainties involved, the estimated ages generally agree with the magnetic anomaly time scale (Fig. 45) when the most reasonable minimum rate (maximum age) is utilized. Station 6 is excluded from Figure 45 because of its position on Sovanco Fracture Zone. The maximum thickness of oxide observed on a specimen from that station is 1 cm, indicating an age less than 4 my. Fission Track Age Dating Specimens from the Mid-Atlantic Ridge ranging in age from 12,000 to 1,300,000 years have been dated by. the fission track technique (Fleischer et al , 1968; Aumento, 1969). However, the uranium contents of the M. A. R. basalts are generally at least an order of magnitude greater than those of basalts from the study area (Table X). Of the twenty-one specimens which were irradiated, only two contain sufficient uranium to make age determinations by counting fission tracks feasible. The ages and relevant data are presented in Table XI. The large confidence intervals associated with the ages are due 166 TABLE X. Minimum U contents of basalts from the study area, determined from fission track concentrations in irradiated specimens. Also shown for comparison are U contents of representative basalts from the Mid-Atlantic Ridge near 45°N (from Aumento, 1969) Dredge Station Feature Specimen Number Minimum U Concentration (ppm) Juan de Fuca Ridge 1 Middle Ridge 70-16-2-4 (1-8) 70-16-2-19 (1-4) 0.006 0.007 2 Middle Ridge 71-23-1-7* (2-2) 0.084 3 West Ridge 71-23-3-1 (3-2) 0.002 4 Endeavour Seamount Sovanco Ridge 71-15-8-1 (4-1) 0.013 5 East End 71-15-5-23 71-15-5-25* 0.011 0.029 6 Southern Flank Heck Seamount Chain 71-23-7-4 (6-8) 71-23-7-8 (6-4) 0.009 0.006 7 East Peak 70-16-5-4 70-16-5-6 70-16-5-23 0.001 0.001 0.002 8 Heck Seamount PZ69-1-7 PZ69-1-21 0.009 0.008 9 Heck Seamount PZ69-2-3 (9-1) PZ69-2-8 0.005 0.008 10 Heck Seamount PZ69-3-2 PZ69-3-3 0.009 0.008 11 West Peak Heckle Seamount Chain 70-16-9-3 70-16-9-7 0.006 0.004 12 East End 71-15-10-7 (12-2) 0.010 Axis of Mid-Atlantic Ridge near 45°N AG-19-66-56-1 AG-19-66-S6-4 CHAIN-43-103 AG-22-68-197 0.25 0.38 0.24 0.09 *Age determined TABLE XI. Fission track data for two specimens from Juan de Fuca and Sovanco Ridges Specimen Number Spontaneous Fission Tracks Observed Area Surveyed (cm2) Spontaneous FT Density (FT/cm2) Induced FT Density* (FT/cm2) Fission Track Age (yrs) 95% Confidence Limits on Age (yrs) 2-2 13 1.04 12.5 38,580 954,000 418,000-1,490,000 (71-23-1-7) ± 28% ± 3.2% 5-12 0 2.17 < 0.46 13,193 < 103,000 0- <309,000 (71-15-5-25) ± 100% ±3.2% *For a neutron dose of 4.81 x 101D ± 3.1% n.v.t. 168 to the small number of spontaneous fission tracks which were found. In specimen 5-12, no spontaneous tracks were found, but the calculations were made as i f one fission track had been observed; hence, the calculated age is a maximum age. a. Specimen 2-2 As described in the section on Petrography of the Basalts, the basalts from Station 2 and nearby Station 1 on Middle Ridge are altered compared to those from West Ridge and Endeavour Seamount. This suggests that the former rocks may be older. Furthermore, some specimens from Station 1 have oxide coatings up to 1 mm in thickness (Fig. 45), again indicative of older rocks, although no measurable oxide coatings were observed on the specimens from Station 2. The interpretation of the magnetic anomalies in terms of the reversal time scale is equivocal, but the most apparent correlation is that the positive anomaly associated with most of Middle Ridge is part of the central Brunhes anomaly (less than 0.7 my), so well-defined along the crest of Juan de Fuca Ridge to the south (Fig. 3 and 4). Both Stations 1 and 2 are on the part of Middle Ridge northeast of the positive anomaly, and are definitely from an area of negative magnetic anomaly (Profile F-F'--Plate VI). Hence, on the basis of the magnetic anomalies, the age of the basalts from these dredge hauls would be over 0.7 my, but certainly no greater than 0.9 my. This agrees reasonably well with the age of 953,000 ± 534,000 years determined for specimen 2-2 by the fission track technique. b. Specimen 5-12 The maximum age of 309,000 years determined for this basalt is not anticipated by the magnetic anomalies because Station 5 is located at the 169 eastern end of Sovanco Ridge in an area of negative anomaly (Fig. 4; Profile G-H--Plate VII). However, parts of Sovanco Fracture Zone do have positive magnetic anomalies associated with them and Sovanco could be a "leaky" transform fault (Menard and Atwater, 1969) caused by changing spreading directions of Juan de Fuca and Explorer Ridges. If so, younger volcanic activity may have been superimposed on older sea floor, in insufficient quantity to change the magnetic sign. The fine-grained basalt represented by specimen 5-12 is the least altered of the basalts recovered at Station 5, and the diversity of the rocks collected, as well as the location of the station, have led to the conclusion that a talus deposit was sampled. Hence, a mixture of older and younger basalt may have been dredged. Implications of the Relative Ages of the Basalts The increasing thicknesses of ferro-manganese coatings with increasing distance from the crest of Juan de Fuca Ridge agree with the increasing magnetic anomaly ages. They also support the suggestion that con-struction of the seamounts comprising the Heck and Heckle chains occurs largely near the crest of Juan de Fuca Ridge (see also p. 89). Ferro-manganese geochronology and the age of specimen 2-2 further support the postulation based on seismic profiles magnetic data, and petrography that West Ridge and Endeavour Seamount (as well as adjacent Endeavour Trough) are loci of most recent volcanic activity on Juan de Fuca Ridge in the study area. Middle Ridge is apparently an older feature in terms of volcanic activity. The young fission track age of specimen 5-12 and the thick ferr-manganese coatings on some specimens from Station 6 do not aid in the 170 interpretation of the complex Sovanco Fracture Zone region. The coatings support the magnetic evidence that most of Sovanco Ridge may be an older feature, formed during a reversal (>0.7 million years ago). However, the young age of specimen 5-12 supports the interpretation that the positive magnetic anomalies on Sovanco Ridge represent magmatic activity during the present Brunhes Magnetic Epoch (Vine, 1968). SEDIMENTARY DEPOSITS Introduction The seismic reflection profiles indicate that igneous basement is covered by sedimentary deposits of varying thickness over much of the study area. Even when dredging over a site of apparent basement outcrop, sediments were often recovered (Tables XIII, XIV, Appendix B). On the continental slope, samples of mudstone and siltstone were recovered, as well as unconso-lidated sediments. Locations of the four stations at which these consolidated deposits were recovered are shown in Figure 29. Because of the lack of fine resolution on the seismic profiles and the inexact method of sampling, the stratigraphic significance of the sedimentary deposits is not known. Sampling by coring or drilling would be necessary in order to establish that. Nevertheless, microfaunal studies to determine the li f e environments and ages of benthonic fauna in the deposits were in some cases able to provide supporting data for interpretations based on the seismic reflection profiles. Except for the microfaunal studies of thirteen samples of uncon-solidated sediment and two samples of mudstone, no detailed studies have been made of the sedimentary deposits. The microfaunal studies were done by B. E. B. Cameron of the Geological Survey of Canada (Vancouver) on samples 171 selected and submitted to him by the author. The results, based on written and oral communications from B. E. B. Cameron (1972), are summarized here. It is obvious from these preliminary data that much valuable information would be obtained from detailed microfaunal and sedimentological studies of both consolidated and unconsolidated sedimentary deposits, particularly i f samples were obtained by coring or drilling. Juan de Fuca Ridge and the Seamount Chains Semi-consolidated Deposits Green slabs of semi-consolidated deposits consisting of foraminifers loosely cemented by ferruginous material were recovered from Station 9 on Heck Seamount. In thin section chips of sideromelane and palagonite are also observed, and the cement includes altered green palagonite, as well as opaque or translucent ferruginous and manganiferous material (specimen 9-4). The chambers of foraminifers are often f i l l e d or lined by the latter two deposits. -Unconsolidated Deposits Five samples from dredge hauls on Juan de Fuca Ridge contain benthonic fauna of Recent age which show no vertical displacements from their l i f e environments. However, a sample from Station 1 on Middle Ridge (Fig. 28) contains two Recent benthonic populations, one "in situ" at about 1850 m and one definitely transported from a shallower depth. No appreciably shallower depths occur on Middle Ridge or in adjacent regions. Hence, the latter fauna may have been transported from the continental slope by turbidity currents and deposited at abyssal plain depths prior to the uplift of Middle Ridge, which would suggest that at least part of the uplift indicated by seismic profiling studies is of Recent age. On the basis of the echo sounder record of the dredging operation, the greatest depth from which the mud sample of 172 Station 1 could have been dredged is 2200 m, which is 300 m above the present level of abyssal plain sedimentation. This suggests at least 300 m of Recent uplift. Alternatively, the fauna could have been transported and deposited as hemipelagic sediment on Middle Ridge above the level of the abyssal plain. More detailed sampling, particularly by coring, and microfaunal and sedimen-tological studies would be required to enable a definite interpretation. The benthonic fauna in a mud sample from Station 7 at the eastern end of the Heck Seamount Chain (Fig. 28) are of Recent age and "in situ" as far as water depth is concerned. Foraminifers comprise more than 30% of the sample, which is a "foram ooze", but the mud also contains a high concentra-tion of glass fragments, as would be anticipated in a young seamount environ-ment . Continental Shelf Unconsolidated Deposits Six samples of mud from the upper continental slope contain fauna of Pleistocene to Recent age. Three of the six were dredged from slightly deeper than the benthonic fauna indicate, and probably reflect the sweeping of material downward from the uppermost slope or the outer shelf. Consolidated Deposits Mudstone dredged from a stratified sequence exposed on the flank of a canyon on the upper continental slope (Station 14, Fig. 29; Profile R-R'--Plate XV) is fragile, pale brownish-grey (when dry), and usually riddled with worm burrows. It is similar to the Late Pliocene mudstone on the continental shelf in this region, and is probably also Late Pliocene in age (B. E. B. Cameron, written communication). It was dredged from slightly shallower water than the fauna indicate, supporting the theory of uplift postulated for the continental slope in the area. 173 On the basis of forarainiferal studies, the ages of the mudstone and siltstone dredged from a ridge on the lower continental slope (Stations 15, 16, 17—Fig. 29; Profile U-U*--Plate XVIII) are uncertain (B. E. B. Cameron, pers. comm.). The mudstone is brownish-grey and friable, but the siltstone is pale brownish-grey, homogeneous, and well-lithified. A similar siltstone was dredged from the same feature during a U.S.G.S. cruise in 1970 under the direction of E. Silver, and on the basis of faunal assemblage was considered to be Pleistocene in age (W. Barnard, pers. comm., 1970; Silver and Barnard, 1971) . This was suggested to indicate that latest uplift (thrusting) at the base of the slope is Pleistocene or younger, because the siltstone is stratigraphically equivalent to or below rocks involved in younger thrust sheets. Conclusions The unconsolidated sediments dredged from Juan de Fuca Ridge and the seamounts contain benthonic foraminifers of Recent age which, with one exception, indicate no vertical displacement. The sample from Middle Ridge contains both an "in situ" population and a population transported from shallower water. These foraminifers could have been transported from the continental slope as hemipelagic sediment, but i t is also possible that they were transported by turbidity currents, in which case, Recent uplift of Middle Ridge subsequent to their deposition is implied. On the upper continental slope, benthonic foraminifers of Pleistocene to Recent age were identified in unconsolidated sediments, some of which were dredged from deeper than their l i f e environments, indicating the sweeping of material from the uppermost slope or shelf. 174 Mudstone from the upper continental slope was dredged from shallower water than the fauna indicate, supporting the theory of uplift of the continen-tal slope postulated on the basis of the seismic reflection profiles. Similarly, from the lower continental slope a faunal assemblage of Pleistocene age in a siltstone stratigraphically equivalent to or below deposits involved in younger thrust sheets supports the occurrence of Pleistocene or younger thrusting and uplift at the base of the continental slope (Silver and Barnard, 1971). CHAPTER V DISCUSSION: RECENT GEOLOGICAL HISTORY OF THE NORTHERN END OF JUAN DE FUCA RIDGE AND RELATED FEATURES The major geological and geophysical data which have led to the generally accepted conclusion that sea floor spreading is occurring along Explorer, Juan de Fuca, and Gorda Ridges were summarized in Chapter II. Also discussed was the Cenozoic history of the northeastern Pacific Ocean in terms of regional plate tectonics. Atwater (1970) showed that the ridge system in that area is probably a remnant of the East Pacific Rise, that Juan de Fuca Plate is a remnant of the Farallon Plate and that a subduction zone has probably existed west of Washington and Oregon during at least the past 30 million years. The purpose of this chapter is to examine the recent geological events within the study area in view of this regional framework and the data obtained during this investigation. The specific problems to be discussed are: 1. Where is the sea floor spreading centre on Juan de Fuca Ridge in the study area? 2. What is the significance of the occurrence of linear seamount chains on the western flank of Juan de Fuca Ridge in the study area? 3. What has been the role of the Sovanco Fracture Zone? 176 4. When and why did a major episode of uplift and deformation occur in the region of the northern end of Juan de Fuca Ridge? 5. What are the effects of the interaction between the Juan de Fuca and American Plates west of southern Vancouver Island? 1. Spreading Centre The spreading centre in the southern part of the study area is clearly defined on the basis of magnetics, seismic reflection data, and physiography. The magnetic anomaly pattern is symmetrical over the past 6 million years at least (Fig. 4 ), and the spreading rate during the past 0.7 my (indicated by the width of the Brunhes anomaly) has been 5.8 cm/year. This region is similar to Juan de Fuca Ridge to the south of the study area. To the north of Endeavour Trough, however, magnetic anomalies lack the symmetry characteristic of Juan de Fuca Ridge to the south. The broad central anomaly branches, and the locus of most recent sea floor spreading is not obvious as i t is further south (Fig. 4 and 46). Physiography, seismic reflection data, magnetic data, and characteristics of dredged rock samples together indicate that the locus of most recent spreading is offset slightly from the northern end of Endeavour Trough to the western flank of West Ridge and West Valley (Fig. 47). a. Physiography West Ridge and West Valley are the most prominent physiographic features on the crest of Juan de Fuca Ridge in the study area. Both extend from the latitude of the end of Endeavour Trough to Sovanco Fracture Zone. East of these features the ridge lacks physiographic relief, with the excep-tion of the northern end of Middle Ridge. 177 FIGURE 46. Magnetic anomalies of Raff and Mason (1961) super-imposed on bathymetry near the northern end of Juan de Fuca Ridge. Bathymetry from Figure 15. Anomaly ages as interpreted by Vine (1968). M Y 7ZZA r-2 L-4 178 FIGURE 47. Postulated axis of most recent sea floor spreading near the northern end of Juan de Fuca Ridge. Bathymetry from Figure 15. 179 b. Seismic reflection data Seismic reflection profiles confirm the prominence of West Ridge and West Valley indicated by the physiography. Middle Valley, although i t is a major basement feature east of Sovanco Ridge, extends south to only about half-way along West Ridge and then dies out. Furthermore, the presence in places in the valley of at least 1200 m of sediment indicates that i t is a relatively old structure, and that recent extrusion has not occurred there in any detectable quantity. Even i f the age of the floor of Middle Valley is greater than 0.7 my (as indicated by the associated negative magnetic anomaly, Fig. 46), the average sedimentation rate is 1.7 m/1000 years, a very high rate. In the median valley of Gorda Ridge to the south, the sedimentation rate in a thick turbidite section cored in the JOIDES Deep Sea Drilling Project was estimated to be only 0.56 m/1000 years (McManus et a l , 1970). It seems unlikely that the floor of Middle Valley is less than 0.7 my in age. Seismic profiles show no evidence of recent extrusive activity on Middle Ridge or East Ridge, or in East Valley. The latter contains a thick sequence of sediments, and East Ridge and most of Middle Ridge are covered by sediment (Fig. 16). Seismic profiles across Juan de Fuca Ridge immediately south of the southern end of Middle Valley show that sediments (and the underlying basement) have been uplifted and tilted to the east from the crest of West Ridge (Profile C-C--Plate III). The impression is that West Ridge is the crest of Juan de Fuca Ridge. The role of West Valley is not clear, as i t contains sediments which are interpreted to be the older Unit A. Profiles suggest that Unit A has been uplifted and tilted to the west on the basement high west of West Ridge. Whether the perched Unit A deposits on the western flank 180 of West Valley have been down-dropped from the basement high to the west, or whether they have been uplifted from the floor of West Valley is not obvious. The implications of the two alternatives are very different. If the former is the case, then the sea floor west of West Ridge may have been uplifted to the present level of the basement high as part of the regional uplift after deposition of Unit A, and West Valley may have subsequently formed as a graben. If the second alternative is the case, then perhaps the basement high west of West Valley is not the result of the regional uplift, but of the sea floor spreading process, which can form similar structures. Descriptions of median rifts on the Mid-Atlantic Ridge (Van Andel and Bowin, 1968) and on Gorda Ridge (Atwater and Mudie, 1968) are in agreement that the walls are tilted outwards, and composed of blocks, the outer ones of which are higher than those near the median line. On Gorda Ridge, perched sediments occur on the tilted blocks (Fig. 48a). Fowler and Kulm (1970) showed that some of these sediments were originally deposited by turbidity currents on the floor of the median valley, and that subsequent tectonism has lifted the sediments as much as 1000 m up the flanks of the valley in approximately the last million years. Apparently similar features of smaller scale may also form on the East Pacific Rise which lacks a true median r i f t (Larson, 1971). By analogy, i t may be that the perched sediments on the western flank of West Valley have been uplifted. On the basis of available data, either alternative is possible. The occurrence of regional uplift following deposition of Unit A has made the structure in the study area complicated and ambiguous. Nevertheless, the western flank of West Ridge and West Valley clearly form the structural "centre" of Juan de Fuca Ridge at the present time. A diagramatic cross-section illustrates the structure (Fig. 48b). 181 FIGURE 48. a. Topography and bottom character of Escanaba Trough and walls on Gorda Ridge. Bottom character (insert) has been deduced from penetration and side-looking sonar records. Dashed line is path of deep-towed FISH. Depths corrected according to Matthews (1939). From Atwater and Mudie (1968). b. Schematic interpretation of the structure of Juan de Fuca Ridge near its northern end. Western flank of West Ridge and West Valley are interpreted to be the axis of most recent sea floor spreading. 182 c. Magnetic Data West Ridge and West Valley are associated with a broad positive magnetic anomaly. However, Middle Ridge also is associated with a positive anomaly and two interpretations of the magnetic anomaly pattern are possible: 1. The Brunhes (central) magnetic anomaly includes the entire area between West Valley and Middle Ridge, and the intervening negative anomaly is a secondary, rather than primary, magnetic feature. 2. Both the West Valley-West Ridge and the Middle Ridge positive lobes represent the Brunhes anomaly, but the intervening negative anomaly is a primary, and hence, greater than 0.7 my in age. This requires that sea floor spreading (basalt intrusion and extrusion) has occurred along two different axes at times during the past 0.7 my. This explanation has also been suggested for the similar split in the Brunhes anomaly near the northern end of Explorer Ridge (A. G. Thomlinson, pers. comm., 1972). The second alternative is favoured because the thick section of sediments in Middle Valley seems to require an old sea floor beneath the valley. If the floor had formed during Brunhes time, i t would probably be not only less than 0.7 but less than 0.4 my in age, since i t occupies a medial position in the anomaly band. This would require what would appear to be an impossibly rapid rate of sedimentation in Middle Valley (more than 3 m/1000 years). Furthermore, there is no apparent feasible mechanism whereby the negative anomaly of more than 300 gammas could be produced as a secondary effect. It cannot be a topographic effect as the negative anomaly extends about 30 km south of the end of Middle Valley. This also renders heating and metamorphism (Watkins and Richardson, 1968) unlikely causes, at least under the southern extension of the negative anomaly. 183 If this alternative is accepted, then the implication is that spreading is no longer occurring in the Middle Ridge area, and that West Ridge-West Valley are the loci of most recent activity. d. Characteristics of Dredged Basalts All the basalts dredged from the study area have compositions characteristic of ocean ridge basalts. However, the basalts from West Ridge (and Endeavour Seamount) are very fresh, and have no ferro-manganese oxide coatings, in contrast to those from Middle Ridge, which show various alteration and weathering effects, and some of which have acquired oxide coatings up to 1 mm in thickness. Furthermore, an age of 954,000 ± 534,000 years was determined by the fission track technique for a specimen from the northern end of Middle Ridge, just northeast of the positive magnetic anomaly. Hence, the conclusion is that the western flank of West Ridge has been the locus of the more recent volcanic activity. Recent activity has apparently also occurred on Endeavour Seamount, supporting evidence for its position near a spreading centre, or perhaps on a small offset between spreading centres. In conclusion, the various data suggest that the most likely loci of sea floor spreading in the study area are Endeavour Trough, and the western flank of West Ridge and West Valley, with a possible slight offset between them (Fig. 47). This does not imply that Middle Valley could not have been the location of the spreading centre prior to 0.7 my ago. The lack of sym-metrical anomalies on the western flank of Juan de Fuca Ridge is a problem, and the origin of the large positive anomaly which replaces them is equivocal. Discussion of this region is included in the section on Sovanco Fracture Zone (p. 186). 184 The northern termination of Juan de Fuca Ridge is not well-defined. West Ridge and West Valley are cut off by the Sovanco Fracture Zone struc-tures, but Middle Valley is cut off by a fault at about the latitude of the northern flank of Sovanco, and the western bounding fault of the valley cuts across the Sovanco Fracture Zone structure (Fig. 16). The basement under Middle Ridge dips north below onlapping sediments from Cascadia Basin and does not seem to be cut by any transverse structures, unless the possibly transverse faults observed at latitude 48°46' represent the Sovanco trend. East Valley and East Ridge seem to terminate at approximately that same latitude. It is thought that the regional uplift has strongly affected the primary structural trends in the northern part of the study area, and further discussion is included in the section concerning that uplift (p. 191). 2. Seamount Chains Menard (1969) postulated that volcanic seamounts remain active as they drift with the crust away from an ocean ridge crest. In contrast, the data available from Heck and Heckle Seamount Chains suggest that those seamounts grew to virtually their present sizes while s t i l l on the crest of Juan de Fuca Ridge: a. The linear magnetic anomalies cross the seamount chains unchanged in several places (Fig. 46). In some cases there is embayment of an older anomaly by the next younger anomaly, indicating that volcanism continued following the polarity reversal and for a short distance from the ridge crest. b. The peaks on the chains have magnetic "highs" or "lows" associated with them of the same sign as the enclosing linear anomaly, indica-ting that they grew during one polarity event (Raff and Mason, 1961). 185 c. No evidence of recent volcanic activity was obtained from the dredge hauls on the seamount chains near their western ends. Many rocks recovered from Stations 11, 12 and 13 (Fig. 28) have ferro-manganese oxide coatings up to 1 cm in thickness, and a thick slab of ferro-manganese pavement was recovered at Station 13 (Fig. 44d). In another haul (Table XIV; 71-23-6) no rocks were recovered, but black manganese smears were observed on the dredge, indicating that oxide pavement may have been deposited over much of the area. d. The chemical composition of fresh seamount chain basalt is very similar to that of ridge basalt. No basalts of alkali affinities were analyzed from the seamount chains. Hence, these basalts could originate from the same source as the ridge basalt. Hence, i t appears that exceptionally large volumes of magma are dis-charged at certain loci on the western side of the crest of Juan de Fuca Ridge. Such discharge is somewhat episodic as discrete peaks are generally formed, and the Heckle locus has been comparatively unproductive in the last 1.5 my. The reason for this localization of magma production is not known, nor is the reason for the lack of similar activity on the eastern flank of Juan de Fuca Ridge. The latter may be somehow related to compression within Juan de Fuca Plate because it is small and caught between two larger plates. If the foregoing interpretation is correct, then the trend of the seamount chains reveals the direction of motion of the Pacific Plate during the past 3.5 my. The chains are not perpendicular to Juan de Fuca Ridge, but intersect the ridge at angles of 15-20°. The fact that the chains are not exactly parallel to one another indicates that the measurement is only approxi-mate. However, this trend is more compatible with the trend of Blanco 186 Fracture Zone, and indicates that the latter may be a better approximation of the spreading direction than the perpendicular to the ridge (Fig. 50a, p. 197). 3. Sovanco Fracture Zone Physiography, seismic reflection data, and magnetics a l l indicate that Sovanco Fracture Zone is not a simple transform fault. It is a broad zone of irregular topography and irregular magnetic anomaly pattern, and contrasts with the narrow trace of the Blanco transform fault at the southern end of Juan de Fuca Ridge (Fig. 3). To explain the complicated magnetic anomaly pattern, Vine (1966, 1968) deduced the presence of two parallel faults in the region between Explorer and Juan de Fuca Ridges (Fig. 3). These have been interpreted to have originated as strike-slip faults (Pavoni, 1966; Peter and Lattimore, 1969). A more detailed discussion of the peculiarities of the Sovanco Fracture Zone region is offered here. Approximately 7.5 million years ago the direction of sea floor spreading in the northeast Pacific began to change from east-west through an angle of 15-20° to west northwest-east southeast. At that time and probably until less than 6 million years ago, there was no offset in the spreading centre where Sovanco Fracture Zone is now because the vertically and obliquely lined anomalies in Figure 3 are not offset (Fig. 49d). Offset of anomalies about 5 million years old (the horizontally lined anomalies on Figures 3 and 4) indicates that this is the earliest time at which Sovanco could have existed. This coincides approximately with the time at which the spreading centre on Juan de Fuca Ridge reached its present orientation. However, i t should be noted that Sovanco could have been initiated as a strike-slip fault zone at any time since then, and continued spreading on Explorer and Juan de Fuca Ridge crests would result in transform movement on the fault zone. 187 FIGURE 49. a-c. Postulated mode of adjustment of a spreading centre to a change in spreading direction (from Menard and Atwater, 1968). d-f. Postulated sequence of events in the development of the northern part of Juan de Fuca Ridge in the last 8 million years (drawn schematically). Dotted arrows indicate unstable plate motions. 188 It was suggested by Menard and Atwater (1968) that a change in spreading direction might result in the break up of a ridge (spreading centre) into segments, joined by transform faults (Fig. 49a, b, c). They predicted that the transform fault zone should become parallel to the new spreading direction very early in the readjustment. It is here postulated that this did not occur during the formation of the Sovanco Fracture Zone because that feature originated as a strike-slip fault, not as a transform fault. The strong east-west trend of magnetic anomalies west of West Valley in the study area suggests that there exists an east-west structural trend (Fig. 46). This postulation is strongly supported by the continuation of Explorer Trench south of the end of the Brunhes anomaly on Explorer Ridge to approximately the southern end of the L-shaped anomaly southeast of Explorer Seamount (Fig. 4). It is suggested that this major topographic feature (Explorer Trench) marks the original position of the Explorer Ridge spreading segment subsequent to the break up of a continuous Juan de Fuca Ridge. This conception of the configuration of the ridge-transform fault system about 5 million years ago is illustrated in Figure 49e. Because at the time of its creation the east-west fault zone was not perpendicular to the ridge crest (as the magnetic anomalies after 5 million years ago were not north-south in trend), i t is suggested that i t originated as a strike-slip fault, perhaps as a result of stresses built up by the rotation of the spreading centre. Motion on Sovanco subsequent to its formation would be of a transform type. Morgan (1968) showed that plate motions need not be perpendicular to the axis of the spreading ridge, although they are generally observed to be so, but that transform faults should indicate the direction of relative motion of plates. The combined evidence of the trends of Heck and Heckle Seamount Chains, Blanco Fracture Zone, and 189 the axis of Juan de Fuca Ridge indicates that an east-west orientation for Sovanco Fracture Zone as a transform fault is unstable during the past 5 million years. Whatever the reason for an in i t i a l east-west break, i t is improbable that i t could persist. The present location of transform motion within the broad Sovanco Fracture Zone region is indeterminate. Faulting observed on reflection profiles south of Sovanco Ridge (Fig. 16) could be related to uplift rather than horizontal motions. The trend of these faults is not compatible with spreading direction indicated by Blanco Fracture Zone. The positive magnetic anomaly north of Sovanco Ridge in Cascadia Basin suggests that i t may be a "leaky" transform fault (Menard and Atwater, 1969), and an age of less than 309,000 years determined by the fission track technique for a basalt specimen from the eastern end of Sovanco Ridge indicates that young volcanic activity has occurred. This explanation is incompatible with the directions of motion which seem to require compression across Sovanco Fracture Zone (Fig. 49f). It is postulated that transform motion on Sovanco Fracture Zone is now approximately parallel to that on Blanco Fracture Zone. The change from the original east-west break may have resulted in the wedge-shaped fracture zone between Juan de Fuca and Explorer Ridges, the southern end of the latter (Explorer Trench) being cut off by the change so that i t no longer acted as a spreading centre. Alternatively i t may have been that a second strike-slip fault trending nearly east-west developed at the southern end of the present Explorer Ridge, as postulated by Vine (1966) (Fig. 3). This would create a broad fracture zone between the ridges, across which transform motion now occurs. 190 In order to explain the lack of symmetry of magnetic anomalies parallel to the inferred spreading direction east and west of Juan de Fuca Ridge near its northern end, i t appears to be necessary to postulate igneous activity in the past 0.7 my to create the east-west anomaly north of Heck Seamount Chain and to destroy older symmetrical anomalies. Seismic reflection profiles show no evidence for such activity. Another problem area is the small "Explorer" sub-plate, north of Juan de Fuca Ridge and east of Explorer Ridge. Its situation is similar to that of the small block east of Gorda Ridge (Gorda Basin) in that the Queen Charlotte and Sovanco fault zones are non-parallel, like the Blanco and Mendocino, and the responses of the two small sub-plates seem to be similar. Both have numerous internal earthquake epicentres (Fig. 5), marked departures from rigid plate tectonics. Internal deformation of Gorda Basin has been described by McEvilly (1968) and Silver (1971b). By means of seismic reflection profiles Silver mapped the fault pattern in the basin, and found northeast-trending faults essentially paralleling the magnetic anomalies indicating deformation along lines of primary weakness. He also found that the Gorda sub-plate may have a component of underthrusting along the Mendocino Fault. Faulting has also been observed on seismic reflection profiles across Explorer sub-plate, as well as basement uplift (McManus et_ a l , 1972). Interaction between the Explorer sub-plate and the sea floor to the south (Sovanco Fracture Zone area and the northern end of Juan de Fuca Ridge) may be responsible for the uplift observed in those regions. This is discussed in the next section which concerns this deformation. In conclusion, insufficient information is available to enable a satisfactory interpretation of the history of the Sovanco Fracture Zone region . Detailed surveys are required of the western part of the fracture 191 zone region, Explorer Trench, and the region north of Sovanco Ridge. Studies of the sprawling Explorer Seamount, located as i t is opposite the inter-section of Sovanco Fracture Zone and Explorer Ridge, might provide information valuable in regional interpretations. 4. Uplift and deformation at the northern end of Juan de Fuca Ridge The distribution of sediments observed on the seismic reflection profiles indicates that the northern end of Juan de Fuca Ridge and at least the eastern half of Sovanco Fracture Zone have been elevated relative to adjacent Cascadia Basin and Juan de Fuca Abyssal Plain, and to Juan de Fuca Ridge to the south. This uplift resulted in faulting and tilting of both igneous basement and the overlying sediments (referred to as Unit A), and the relief of the structures comprising the northern end of Juan de Fuca Ridge was greatly increased. This secondary relief controlled deposition of subse-quent sediments (Unit B), although minor deformation continued throughout the time of deposition of Unit B. The evidence for this uplift and the details of sediment distribution, faulting, and deformation are discussed in Chapter III. It is the purpose of this section to discuss the timing and possible causes of the event. The depth of the sea floor on the northern end of Juan de Fuca Ridge at the end of Unit A time (prior to uplift) must have been approximately the same as that in adjacent Cascadia Basin because Unit A was deposited virtually everywhere, with the exception of West Ridge and a small part of Middle Ridge (Fig. 16). Unit A was also deposited on the southern flank of Sovanco Ridge, on the part of Juan de Fuca Abyssal Plain north of Heck Seamount Chain, and in West Valley. This area was also at uniform depth, although that need not have been the same depth as the Cascadia Basin-Juan de 192 Fuca Ridge level. The latter depth can be estimated from the present depth to Unit A in Cascadia Basin east of East Ridge and north of Juan de Fuca Ridge, where i t is approximately 2700 m below sea level. As turbidites were apparently not deposited on the crest of Juan de Fuca Ridge south of the study area (Lister, 1970) but only on its eastern flank, most of the crest must have been less than 2700 m deep. Depths there now are commonly 2500-2700 m; hence, the elevation of the ridge has apparently not changed except in the study area. By comparing the depth of 2700 m to the present depths to Unit A on various features, rough estimates of the amounts of uplift can be made: East Ridge, 300 m; East Valley, none; Middle Ridge, 550 m (maximum); Middle Valley, none?; south of Middle Valley, 200 m; West Ridge, 450 m. The amounts of uplift of Middle Ridge and East Ridge diminish to the south. The amount of uplift indicated on the eastern flank of Juan de Fuca Ridge east of Endeavour Trough in the southern part of the study area is only 75 m at most. In the region south of Sovanco Ridge i t has been shown that the depth to the top of Unit A was probably originally about 2925 m. Comparisons with the present depths to Unit A on various features in that region record the following amounts of uplift: Sovanco Ridge, 775 m maximum; West Valley, 85 m; basement high west of West Valley, 525 m maximum. On Juan de Fuca Abyssal Plain near Sovanco Fracture Zone, the top of Unit A may have been depressed by more than 100 m after its deposition, although identification of the top of Unit A in this region is uncertain. The sediment distribution indicates that ridges were uplifted but valleys were l i t t l e affected. The deformation apparently did not create new structures but utilized those already in existence. 193 A minimum age for the uplift can be estimated by comparing the thicknesses of Unit A and Unit B in Middle Valley, assuming that turbidity currents have had access to that feature for 0.7 my (probably a valid assumption) and that sedimentation rates there have been constant during that period (probably not valid, but hopefully not extremely inaccurate when averaged over that time span). The assumption of a 0.7 my age for the basement under Middle Valley has already been discussed (p. 182). Although the Unit A-Unit B contact cannot be definitely identified in Middle Valley, i t is tentatively identified at a depth of about 170 m below the sea floor; certainly i t is no shallower than that. Assuming a sedimentation rate of 1.7 m/1000 years, the minimum age of the Unit A-Unit B contact is 100,000 years. If, as suggested below, the rate has not been constant, then the minimum age as calculated above will be less. This approximately coincides with the time of the end of the Sangamonian Interglacial. However, i f the Holocene sedimentation rate of 10 cm/10,000 years which was measured in a core from the median valley on Gorda Ridge (Moore, 1970) is typical of interglacial rates in the northeastern Pacific, then most of Unit A was probably deposited prior to the Sangamonian, that i s , before about 350,000 years ago. Hence, the major uplift and deforma-tion could have occurred over a time interval of 250,000 years. It is concluded that Unit B was deposited primarily during the Wisconsin Glaciation and Unit A primarily during the Illinoian Glaciation. Uplift and deformation occurred during the Sangamonian Interglacial, but i t is improbable that they were related to the glacial cycle. Their coincidence with the Unit A-Unit B contact probably need not imply a direct relationship, although certainly the contact is made obvious by the deformation. It is possible that the deformation may be responsible for part of the difference in appearance of the reflectors within the two units. 194 Obviously the foregoing discussion contains considerable speculation, the applicability of which could be proved or disproved by the study of long cores or d r i l l samples from the region. The problem of the cause of the uplift and deformation s t i l l remains. Two possibilities are discussed here: a. Episodal sea floor spreading, and b. interaction with the Explorer sub-plate to the north. a. Episodal sea floor spreading Changes in the rates of sea floor spreading are probably accompanied by changes in elevation of the ridge crests and flanks (Schneider and Vogt, 1968). It has been shown that during the past 0.7 my, two axes of spreading may have functioned near the northern end of Juan de Fuca Ridge, but the Middle Ridge axis has not been the locus of most recent spreading (see p. 181). The uplift could have been related to the transfer of spreading from Middle Ridge to West Ridge-West Valley and/or an increased rate of spreading. On the Mid-Atlantic Ridge south of the Vema Fracture Zone, recent uplift of 500-2000 m has occurred more than 100 km west of the ridge axis (van Andel, 1969). The evidence is uplift and tilting of originally horizontal Pleistocene turbidites. Van Andel concluded that "uplift of the currently active crest of the Mid-Atlantic Ridge is responsible". That sea floor spreading is episodal has also been documented. Moore (1970) showed that no volcanism has occurred in at least parts of the median valley of Gorda Ridge for 100,000 years. Discontinuities in spreading and changes in spreading rates have been suggested on the basis of magnetic anomalies topography, and sediment thickness in many areas (Schneider and Vogt, 1968; Ewing et a l , 1968; Hays and Ninkovich, 1970). A difficulty in this explanation is why only the northern tip of Juan de Fuca Ridge would be affected. The uplift of 195 Sovanco Fracture Zone could also be related to a component of sea floor spreading in that region (a "leaky" tranform fault), although, as previously discussed, inferred stresses in the region indicate compression, not extension. b. Interaction with Explorer sub-plate to the north Silver (1971b) has shown that Mendocino Escarpment probably has had a component of underthrusting by the Gorda block to the north. Seismic reflection profiles across the escarpment and the ridge behind i t show sub-bottom reflectors paralleling surface morphology which has a dip to the south. The structure indicates gentle uplift along Mendocino Escarpment. An earth-quake first-motion study allows either overthrusting southward or northward. Silver concluded that the structural attitude of sub-bottom reflectors south of the escarpment is consistent with south-side up movement and with the second solution of a south-dipping reverse fault at depth (overthrusting northward). It is here proposed that similar underthrusting of Sovanco Ridge and the northern end of Juan de Fuca Ridge by the Explorer block may be the cause of the uplift which has occurred there. Certainly the structure of Sovanco Fracture Zone is compatible with the underthrusting of a block from the north. The faults on the southern flank may be adjustments to the uplift, or they may be largely strike-slip (transform) faults related directly to spreading on Juan de Fuca and Explorer Ridges. The structure of Juan de Fuca Ridge is more complex; however, that is understandable because the ridge had a primary structural pattern as a result of sea floor spreading. Regional uplift would probably result in faulting along lines of primary weakness. 196 A major problem with this interpretation is the timing of the uplift. It apparently occurred rather abruptly, certainly over a time span of no more than 250,000 years (the length of the Sangamonian Interglacial). It seems likely that underthrusting of the Explorer block would have occurred over the last 5 my. One possibility is that spreading on Explorer Ridge is episodic, and hence, so is the underthrusting. In conclusion, neither of the causes put forth to explain the uplift is very satisfactory. A combination of the two possibilities seems to be necessary--fluctuations in spreading rate or pulses of spreading on both Explorer and Juan de Fuca Ridges (on the latter at the West Ridge-West Valley axis rather than on Middle Ridge) between 100,000 and 350,000 years ago and underthrusting of Sovanco Fracture Zone and Juan de Fuca Ridge by Explorer sub-plate. Continued spreading and minor underthrusting could account for continued minor deformation during deposition of Unit B. 5. The Subduction Zone Construction of the velocity vectors between the Pacific, American, and Juan de Fuca Plates indicates oblique underthrusting of the continental slope by the Juan de Fuca Plate (Fig. 50a). The lack of a Benioff zone of earthquakes east of the Juan de Fuca Plate is evidence against the occurrence of underthrusting at the present time. However, a few earthquakes of intermediate depth do occur (Tobin and Sykes, 1968) and boundaries where compression is predicted to be slow are observed to be less seismic; for example, the South Island of New Zealand and Macquarie Ridge, the Aleutians near 170°E, and southern Chile (LePichon, 1968; Isacks et a l , 1968). The critical temperature above which earthquakes cannot occur is about 800°C. Observations of maximum depths of earthquakes behind trenches 197 FIGURE 50. a. Vectors diagrams for plate motions in the northeast Pacific. Jp = motion of Juan de Fuca Plate relative to Pacific; Ap = motion of American Plate relative to Pacific, and J ^ = motion of Juan de Fuca Plate relative to American. Jp parallel to Blanco Fracture Zone is considered more likely than Jp perpendicular to Juan de Fuca Ridge. b. Schematic block diagram of continental slope showing inter-pretation of structures caused by compressive motion between Juan de Fuca and American Plates. Modified after Silver (1971a). 198 compared with dips of Benioff Zones and rates of underthrusting indicate that it takes about 10 my for an increment of down-going slab to reach this temperature (Isacks et_ al_, 1968) . If the down-going slab east of Juan de Fuca Ridge has a dip of less than 50° (see below), then the maximum depth of earthquakes expected would be only 150 km, i f the rate of underthrusting is only 2.5 cm/year as indicated by the vector construction assuming Juan de Fuca-Pacific Plate motion to be parallel to Blanco Fracture Zone (Fig. 50a), the more likely alternative (see p. 185). This depth is probably even less because, as suggested by Atwater (1970), Juan de Fuca Plate is small and young, and would be expected to be quite thin even at its eastern margin. Hence i t would reheat quickly. The above argument is opposed by a recent detailed study of travel time anomalies of a Seattle earthquake (McKenzie and Julian, 1971). This supports the existence of a high velocity slab dipping 50°E beneath south-western Canada and northwestern United States. Why the slab has so few earth-quakes associated with i t is not understood, but i f present, i t is evidence for subduction within the past 10 my. The line of Quaternary andesitic and " rhyolitic volcanoes (Wilcox, 1965) stretching from southern British Columbia to central California also suggests that underthrusting has occurred in the late Cenozoic (Fig. 5). If the dip of the subduction zone increases from about 10° under the continental slope (see p. 93) to 50° inland, then the depth of derivation of these magmas is about 100-125 km, which is compatible with proposed models (Ringwood, 1969). Although no topographic trench exists east of Juan de Fuca Ridge, seismic reflection profiles show the presence of a sediment-filled trench (Hayes and Ewing, 1970; Hamilton, 1967; McManus et al , 1972; Profiles of the present investigation). Most trenches postulated to be the sites of 199 subduction are essentially devoid of sediment (i.e., Tonga Trench, New Hebrides Trench). However, the eastern Aleutian Trench and the southern portion of the Peru-Chile Trench are f i l l e d , and seismic profiles across those features are similar to profiles across Cascadia Basin (Hayes and Ewing, 1970). All contain ±2 km of turbidite sediments which are essentially undis-torted but gently dipping towards the continent. The turbidite sequences show an increase in thickness of every bed towards the land. This implies the presence of a hinge approximately 100 km seaward of the trench axis. In the case of the Juan de Fuca Plate, this hinge from which the sea floor t i l t s eastward is the crest of Juan de Fuca Ridge. The difference between the sediment-filled and unfilled trenches emphasized by Hayes and Ewing (1970) is the comparative steepness of the landward wall of the f i l l e d trenches. They believe that this is due to the obscuring of the true character of the back wall by highly disturbed sediments. The steep slope af the foot of the back wall is the result of a sharp front of deformation of the unconsolidated trench sediments. This is in agreement with the conclusions of this writer, based on the study of seismic reflection profiles from the continental slope in the study area. It has been argued that the presence of undeformed turbidites in trenches precludes the occurrence of underthrusting (Scholl et a l , 1968). However, the effect that compressive deformation should have on unconsolidated sediment has been the subject of controversy. Evidence for compressive deformation of trench sediments has been presented by Chase and Bunce (1969) for the eastern margin of the Antilles. Silver (1969, 1971a) showed that folding and faulting of sedimentary deposits on the continental slope off northernmost California can best be explained as resulting from oblique underthrusting. Grow (1970) presented evidence for the presence of a ridge 200 composed of complexly deformed and lithified sediment at the base of the northern wall of the Aleutian Trench. Holmes et al_ (1972) present a seismic reflection profile in which turbidites adjacent to the back wall of the Aleutian Trench are deformed by folding. The seismic profiles restrict deformation from plate interaction to a narrow belt along the northern trench wall and possibly below the trench wall. Hence, i t appears that deformation of a trench f i l l occurs only at the leading edge and that deformation is not transmitted very far horizontally. Chase and Bunce (1969) suggested that folding and faulting of trench sediments and slump deposits form an acoustically absorptive layer that hides the zone of active deformation. Furthermore, i f the direction of underthrusting is oblique to the continental margin, as is the case in the study area, the shear component would cause faulting and tend to restrict deformation to a narrow zone. Seismic reflection profiles from the study area indicate that the igneous basement is not involved in the deformation, at least i n i t i a l l y , but the overlying sediments are intensely deformed. A process of continental accretion is occurring, with part of the sediment derived directly from the continent and part uplifted from Cascadia Basin. Sequential folding, faulting, and uplift of Cascadia Basin deposits build the continental slope. The Cascadia Basin sediments do not retain coherent internal reflectors after deformation and act as a basement. Sediments derived directly from the continent drape and i n f i l l this acoustic basement. The process is described in detail in Chapter III. It is similar to that suggested by Silver (1969, 1971a) for the continental margin off northernmost California. He interprets the faults to be l i s t r i c in nature. A model very similar to that presented by Silver (1971a) is postulated for the structure of the continental slope in the study area (Fig. 50b). 201 It was shown in Chapter III that the Raff and Mason magnetic anomalies can be traced under the continental slope for a distance of about 40 km, i.e., up to anomalies nearly 8 million years old. The definite presence of anomalies as young as 5.5 million years under the continental slope is possible evidence for underthrusting during the late Cenozoic. On the basis of increasing anomaly source depths, the basement dips under the continental slope at an angle of 1 0°. The magnetic data from the continental slope in the study area do not confirm the observations of Emilia et al_ (1968) that two of the Raff and Mason anomalies extend to the shelf break with their widths unchanged and their magnitudes only slightly reduced. In the study area the linear magnetic anomalies dramatically increase in wave length and decrease in amplitude under the continental slope compared to their counterparts in Cascadia Basin (Fig. 26) and to the corresponding anomalies south of 48°N (Raff and Mason, 1961). The fact that the source depths of the anomalies of Emilia et^  al_ are very shallow west of Juan de Fuca Strait suggests that the continuity of those anomalies with the Raff and Mason anomalies is fortuitous, and that they are instead caused by the Tertiary volcanics which underlie the shelf in that area. These rocks were encountered in the wells drilled in the area by Shell Oil Limited below Lower Miocene sedimentary rocks at depths as shallow as 2000 m (Shouldice, 1971). A magnetic profile run by the author from Cascadia Basin into Juan de Fuca Strait crossed from the area of low-amplitude, long wave-length anomalies of the slope into an area of virtually no magnetic relief on the upper slope and outer shelf, and then into an area of considerable magnetic relief near the mouth of Juan de Fuca Strait. This was interpreted by the author to be the area underlain by Tertiary volcanics, equivalent to those which occur on southern Vancouver Island and on the Olympic Peninsula. These volcanic rocks are not oceanic basement. 202 It has been argued that the trench east of Juan de Fuca Ridge was abandoned and f i l l e d with sediment after the change in spreading direction on Juan de Fuca Ridge, 7.5 million years ago (LePichon, 1968). Silver (1971a) suggested instead that the present regime may be the result of the slow rate of subduction predicted for the Juan de Fuca Plate combined with the high sedimentation rate in the region, since the change in spreading direction in late Miocene time. The higher rate of subduction predicted for earlier times (Atwater, 1970) would drag the sediment into the mantle. Certainly the deformation at the base of the continental slope has involved a l l but the most recent deposits in Cascadia Basin (Profiles T-T1 and U-U'—Plates XVII and XVIII). A Pleistocene age for siltstone dredged from a ridge composed of deformed deposits near the base of the slope (Silver and Barnard, 1971) indicates that deformation has occurred in the last 2 million years. Compressive structures on the continental shelf and micropalaeonto-logical evidence of uplift of deep water faunas suggests that similar processes may have occurred during much of the Cenozoic era west of southern Vancouver Island (Tiffin et a l , 1972). CHAPTER VI SUMMARY AND CONCLUSIONS The six major physiographic features in the study area are Juan de Fuca Ridge, Sovanco Fracture Zone, the seamount chains, Juan de Fuca Abyssal Plain, Cascadia Basin, and the continental slope. The structures of these features, determined from bathymetry and seismic reflection profiles, are summarized below. 1. Juan de Fuca Ridge in the northern and central part of the study area consists of three NNE-trending basement ridges (East, Middle, and West), the intervening valleys, and a broad area of uplifted basement west of West Valley. West Ridge and West Valley form the structural centre of Juan de Fuca Ridge north of 48°20' and on the basis of bathymetry, seismic profiles, magnetic data, and dredged basalt samples are postulated to form the axis of most recent sea floor spreading there. South of 48°20', the spreading centre is offset 10 km to the east along a dextral transform fault to Endeavour Trough, a broad basement depression, devoid of sediment, whose eastern margin, at least in places, is a fault. The valleys and ridges east of West Ridge are most prominent near their northern ends and decrease in relief and eventually die out to the south. The ridges consist of one (in the north) or more fault blocks, generally east-tilted and capped by turbidites (Unit A). The latter were originally flat-lying but have been faulted, uplifted, and tilted with the underlying basement. East and Middle Valleys are generally fault-bounded and contain thick 204 sequences of sediments, including both Unit A (originally continuous and co-planar with Unit A on the ridges) and overlying Unit B, also turbidites, whose deposition occurred only in the valleys. Middle Valley, at its northern end where i t lies between Middle Ridge and Sovanco Fracture Zone, is separated from Cascadia Basin to the north by a fault cutting both igneous basement and Unit A. 2. The topographic expression of Sovanco Fracture Zone is a broad, hummocky basement ridge (Sovanco Ridge) trending WNW. The southern flank of this ridge consists of a series of faults which indicate vertical movement but which may also be loci of horizontal displacements. Much of the positive magnetic anomaly associated with Sovanco is north of the ridge itself but of approximately the same trend. A positive anomaly trending east-west south of Sovanco Ridge may represent recent activity along an old Sovanco Fracture Zone trend. Turbidites equivalent to Unit A on Juan de Fuca Ridge have been faulted and uplifted with the underlying basement on the southern flank of Sovanco Ridge. The northern boundary of Sovanco Ridge, at least in places, is a fault or series of faults, poorly developed compared to those on the southern flank. At its eastern end Sovanco Ridge merges into West Ridge. 3. NW-trending Heck and Heckle Seamount Chains consist of series of volcanic peaks coalescing at their bases. Linear magnetic anomalies cross the chains with only minor distortion. 4. Turbidite sedimentation over an irregular igneous basement has smoothed the sea floor on Juan de Fuca Abyssal Plain to a uniform depth of about 2780 m (uncorrected). North of Heck Seamount Chain both Unit A and the younger Unit B have been deposited, but over most of the abyssal plain where the basement is less than 2925 m deep only Unit B has been deposited. 205 5. Cascadia Basin east of Juan de Fuca Ridge is a sediment-filled trench containing more than 2 km of sediment near the base of the continental slope. The igneous basement which dips east from the crest of Juan de Fuca Ridge at about 0.5° has a comparatively smooth surface with no seamounts as on the western flank. Units A and B can be positively identified only near the ridge crest. North of the ridge, Cascadia Basin has a gently undulating basement surface overlain by Units A and B. 6. The continental slope consists of blocks of acoustical basement draped and infilled by stratified sediments. The blocks are composed of sedimentary material (confirmed by dredging) interpreted to have been originally deposited in Cascadia Basin, and subsequently folded and uplifted along reverse faults. Stages in the development of these structures are observed on seismic reflection profiles. The acoustical character of the sedimentary basement results from deformation of internal reflectors. Over-lying stratified sediments were derived directly from the continent. Basalt was dredged from thirteen locations on Juan de Fuca Ridge, Sovanco Ridge, and Heck and Heckle Seamount Chains, indicating that the igneous basement observed on the seismic profiles is largely basaltic. a. Basalt from Juan de Fuca and Sovanco Ridges included glassy pillow and flow fragments, and holocrystalline specimens. Many of the glassy specimens are partial cumulates, containing 10-20% plagioclase (An85-88) or olivine (Fo90-92) xenocrysts. Holocrystalline basalts were dredged from two stations both on fault scarps which may have exposed interiors of flows or intrusions. The youngest basalts, on the basis of their freshness and lack of ferro-manganese oxide coatings, were dredged from West Ridge and Endeavour Seamount. Specimens from Middle Ridge are generally more weathered, have 206 ferro-manganese oxide coatings to 1 mm in thickness, and for one specimen an age of 0.95 ± 0.53 m.y. was determined by the fission track method. This age is in reasonable agreement with the age of 0.7-0.9 m.y. predicted by the magnetic anomaly time scale. A specimen from the eastern end of Sovanco Ridge yielded a fission track age of <309,000 years. b. The seamount chain basalts also include pillow fragments but irregular and glassy flow fragments are more common. Tuffaceous material is also common, implying that seamount eruptions were more violent in nature than the ridge eruptions which is consistent with the expected difference between central vent and fissure eruptions. The lack of highly xenocrystic specimens among the seamount chain basalts suggests that their magmas may reach the surface more quickly than the typical ridge basalt. Increasing degrees of weathering and thicknesses of ferro-manganese oxide coatings on dredged basalts suggest that the seamounts increase in age away from the crest of Juan de Fuca Ridge. c. Twenty-seven representative samples of basalts from Juan de Fuca Ridge, Sovanco Ridge, and Heck and Heckle Seamount Chains were determined to be low-potassium tholeiites of the ocean ridge type. No alkali basalts typical of seamounts in other submarine areas were dredged. The analyses are similar, and indicate no well-developed differentiation trend. The analyzed specimens cluster around the boundary between the olivine- and quartz-normative fields (assuming a constant ratio of Fe3 + to Fe2 + of 0.1 to calculate norms). Comparisons with experimental results on the origin of basalt indicate that the basalts in the study area were derived by essentially direct segregation from the mantle at a depth of about 15 km, and generally reached the surface with only minor modification by fractionation to form partial olivine or plagioclase cumulates. 207 d. Fifteen analyses of altered basalts and of basaltic glass showed that in many cases major chemical changes have occurred. The most common type of alteration—the formation of a dark rim around a specimen—is apparently caused by submarine weathering. The chemical changes which result (notably loss of magnesia and gain of iron and potash) could be mistaken for effects of differentiation. The origin of an alumina-enrichment trend observed in glass of non-pillowed basalts and in altered basalts from Heck Seamount Chain is not understood. Various aspects of the recent geological history of the study area can be interpreted on the basis of physiography, seismic reflection and magnetic data, and dredged samples. i . During the period of reversed polarity (0.7-0.9 m.y. ago) preceding the Brunhes Geomagnetic Polarity Epoch the spreading centre near the northern end of Juan de Fuca Ridge may have been along Middle Valley and the sea floor immediately to the south, where there is a negative magnetic anomaly. Early in the Brunhes Geomagnetic Polarity Epoch (0.7 m.y. ago), and prior to the regional uplift, Middle Ridge may have functioned as a spreading centre because the eastern branch of the Brunhes anomaly is associated with the ridge. However, more prolonged and recent spreading has occurred on an axis along West Ridge-West Valley with which the western branch of the Brunhes anomaly is associated. i i . A widespread tectonic event occurred 100,000-350,000 years ago which increased the structural complexity of the northern end of Juan de Fuca Ridge and Sovanco Fracture Zone, and caused the anomalous relief of the northern end of Juan de Fuca Ridge compared to the ridge to the south. Prior to this event the primary structural features parallel to the magnetic anomalies on 208 Juan de Fuca Ridge were present, but the relief was subdued. Turbidites were deposited over most of the area except for a small part of Middle Ridge, West Ridge (perhaps because much of the sea floor there may have formed after the tectonic event), peaks on Sovanco Ridge, and the seamount chains. The depth to the surface of this turbidite "plain" was about 2700 m on Juan de Fuca Ridge but south of Sovanco Ridge and west of West Ridge i t was deeper—about 2925 m. The tectonic event involved uplifts of as much as 550 m on Juan de Fuca Ridge and 775 m on Sovanco Ridge. Unit A was faulted, uplifted, and tilted with the basement. Block faulting followed the primary trends, and pre-existing ridges were uplifted but valleys l i t t l e affected. Subsequent Unit B deposition was controlled by this secondary relief and hence confined to the valleys. The direct effects of sea floor spreading are difficult to separate from the effects of this regional event; for example, the uplift of basement and Unit A west of West Valley could be either a direct result of spreading or of regional uplift. The regional uplift may have been related to an increase in rate of spreading on Juan de Fuca Ridge, the change in axis of spreading from Middle Ridge to West Ridge-West Valley, compressional inter-action between Explorer sub-plate to the north and Sovanco Fracture Zone and the northern end of Juan de Fuca Ridge, or combinations of these. i i i . Sovanco Fracture Zone may have originated about 5 m.y. ago as a strike-slip fault, or fault zone, trending east-west. This configuration may have been related to the rotation of the axis of the previously continuous and north-trending Juan de Fuca Ridge to a NNE orientation, which was completed about that time. Recent (<0.7 m.y.) igneous activity along the old east-west Sovanco trend has resulted in a positive anomaly of that trend south of Sovanco Ridge, and the consequent disruption of NNE-trending magnetic anomalies in that area. 209 iv. The seamounts comprising the Heck and Heckle Seamount Chains originated and grew to virtually their present sizes while at or near the western crestal area of Juan de Fuca Ridge. This required rapid magma production, which may explain why no xenocrystic specimens are present. v. The presence of magnetic anomalies 5-8 m.y. in age beneath the continental slope is possible evidence for underthrusting of the continental slope by Juan de Fuca Plate. Because all but the most recent deposits in Cascadia Basin are in some places involved in the deformation at the base of the slope, i t is suggested that the relative motion continued until at least Late Pleistocene. It is concluded that Juan de Fuca Ridge near its northern end is an active centre of sea floor spreading where recent extrusion of low-potassium tholeiitic basalt typical of ocean ridges has occurred. However, many of the physiographic features of the northern end of Juan de Fuca Ridge owe their prominence not directly to sea floor spreading but to tectonic uplift which occurred 100,000-350,000 years ago. Underthrusting of the American Plate by the Juan de Fuca Plate has occurred adjacent to the northern end of Juan de Fuca Ridge until at least Late. Pleistocene. 210 RECOMMENDATIONS FOR FURTHER WORK 1. Coring or preferably drilling on the lower and middle continental slope (in an area where magnetic basement age can be predicted by the magnetic anomaly time scale) to determine origin and age of sedimentary material composing the slope structures. 2. Coring or preferably drilling on the crest of Juan de Fuca Ridge to determine stratigraphy and age of sediments overlying igneous basement and to compare Units A and B identified by seismic reflection profiling. 3. A detailed survey of Sovanco Fracture Zone, particularly its western end from which few data are presently available. The survey should include seismic reflection and magnetic profiling, dredging, and coring to attempt to determine the structure and geologic history of this complicated area. 4 . Detailed surveys of Explorer Trench and Explorer Seamount, including seismic reflection and magnetic profiling, as well as dredging on the seamount and the ridges around Explorer Trench, and coring in the Trench. The object would be to determine the origins and ages of these features, and to relate them to the development of Juan de Fuca and Explorer Ridges. 5. Further studies of samples already dredged from the northern end of Juan de Fuca Ridge, Sovanco Ridge, and Heck and Heckle Seamount Chains. For example: comparisons of mineralogy, chemical composition, and rates of accumulation of various types of Fe-Mn deposits (pavements, coatings, nodules, and iron deposits); trace element, rare earth element, and isotope studies of basalts. 211 BIBLIOGRAPHY Atwater, T. 1970. Implications of plate tectonics for the Cenozoic tectonic evolution of Western North America. Bull. Geol. Soc. Am. 8_1: 3513-3536. Atwater, T. M., and J. D. Mudie. 1968. Block faulting on the Gorda Rise. Science 159: 729-731. Aumento, F. 1967. Magmatic evolution on the Mid-Atlantic Ridge. Earth Planet. Sci. Lett. 2: 225-230. 1968. The Mid-Atlantic Ridge near 45°N. II. Basalts from the area of Confederation Peak. Can. J. Earth Sci. 5_: 1-21. 1969. The Mid-Atlantic Ridge near 45°N. V. Fission track and ferro-manganese geochronology. Can. J. Earth Sci. _6: 1431-1440. Aumento, F., D. E. Lawrence, and A. G. Plate. 1968. The ferro-manganese pavement on San Pablo Seamount. Can. Geol. Surv. Paper 68-32, 30 pp. Aumento, F.., and B. D. Loncarevic. 1969. The Mid-Atlantic Ridge near 45°N. III. Bald Mountain. Can. J. Earth Sci. (>: 11-23. Aumento, F., B. D. Loncarevic, and D. I. Ross. 1971. Hudson geotraverse: Geology of the Mid-Atlantic Ridge at 45°N. Phil. Trans. Roy. Soc. Lond. A. 268: 623-650. Bailey, E. H., W. P. Irwin, and D. L. Jones. 1964. Franciscan and related rocks, and their significance in the geology of western California. Cal. Div. Mines Geol. Bull. 183, 177 pp. Barnes, S. S., and J. R. Dymond. 1967. Rates of accumulation of ferro-manganese nodules. Nature 213: 1218-1219. Barr, S. M., and R. L. Chase. 1971. Structure of the continental slope west of Vancouver Island between 48°N and 49°10'N. Can. Geol. Surv. Paper 71-1: 249-251. Bass, M. N. 1971. Variable abyssal basalt populations and their relation to sea floor spreading rates. Earth Planet. Sci. Lett. 11_: 18-22. Bender, M. L., T. L. Ku, and W. S. Broecker. 1970. Accumulation rates of manganese in pelagic sediments and nodules. Earth Planet. Sci. Lett. 8_: 143-148. Bertrand, W. G. 1972. A geological reconnaissance of the Dellwood Seamount area, northeast Pacific Ocean, and its relationship to plate tectonics. M.Sc. dissertation, University of British Columbia, 151 pp. 212 Bonatti, E. 1970. Deep sea volcanism. Naturwissenschaften 5_7_: 379-384. Byrne, J . W., G. A. Fowler, and J. J . Maloney. 1966. Uplift of the continental margin and possible continental accretion off Oregon. Science 154: 1654-1656. Cann, J. R. 1969. Spilites from the Carlsberg Ridge, Indian Ocean. J. Petrol. 1£: 1-19. Cann, J . R., and F. J . Vine. 1966. An area on the crest of the Carlsberg Ridge; petrology and magnetic survey. Phil. Trans. Roy. Soc. Lond. A. 259: 198-217. Carlisle, D. 1963. Pillow breccias and their aquagene tuffs, Quadra Island, British Columbia. J. Geol. 71_: 48-71. Carson, B., and D. A. McManus. 1971. Analysis of turbidite correlation in Cascadia Basin, Northeast Pacific Ocean. Deep-Sea Res. 1_8: 593-604. Carter, L. 1971. Surficial sediments of Barkley Sound and adjacent continental shelf, Vancouver Island, British Columbia. Ph.D. dissertation, University of British Columbia, 196 pp. Chase, R. L., and E. T. Bunce. 1969. Underthrusting of the eastern margin of the Antilles by the floor of the western North Atlantic Ocean, and origin of the Barbados ridge. J. Geophys. Res. 74: 1413-1420. Chase, R. L., S. Barr, and A. G. Thomlinson. 1970. Deformation revealed by CSP on margins of the northern sector of Juan de Fuca Lithospheric Plate, Northeast Pacific Ocean. Geol. Soc. Am., Abstract with Programs for 1970, 2: 518. Chase, R. L., and D. L. T i f f i n . 1972. Queen Charlotte Fault-zone, British Columbia. Intl. Geol. Congress, Proc. of 24th Session, Montreal, Section 8, 17-27. Chase, T. E., H. W. Menard, and J. Mammerickx. 1970. Bathymetry of the North Pacific, Chart No. 4. Scripps Institution of Oceanography and Institute of Marine Resources. Couch, R. W. 1969. Gravity and structures of the crust and subcrust west of Washington and British Columbia. Ph.D. dissertation, Oregon State University, Corvallis. Cronan, D. S., and J. S. Tooms. 1969. The geochemistry of manganese nodules and associated pelagic deposits from the Pacific and Indian Oceans. Deep-Sea Res. 1_6: 335-359. Dehlinger, P., R. W. Couch, and M. Gemperle. 1968. Continental and oceanic structure from the Oregon coast westward across the Juan de Fuca Ridge. Can. J . Earth Sci. 5: 1079-1090. 213 Dehlinger, P., R. W. Couch, D. A. McManus, and M. Gemperle. 1970. Northeast Pacific Structure in The Sea, A. E. Maxwell, Ed., Part II, 133-189, Wiley-Interscience, New York. Dietz, R. S. 1963. Collapsing continental rises: an actualistic concept of geosynclines and mountain building. J. Geol. 71: 314-333. Dobrin, M. B. 1960. Introduction to geophysical prospecting. McGraw-Hill, New York, 446 pp. Drever, H. I., and R. Johnston. 1957. Crystal growth of forsteritic olivine in magmas and melts. Trans. Roy. Soc. Edin. 63: 289-311. Duncan, J . R., Jr. 1968. Late Pleistocene and postglacial sedimentation and stratigraphy of deep-sea environments off Oregon. Ph.D. dissertation, Oregon State University, Corvallis, 222 pp. Emelia, D. A., J . W. Berg, Jr., and W. E. Bales. 1968. Magnetic anomalies off the northwest coast of the United States. Bull. Geol. Soc. Am. 7£: 1053-1062. Engel, A. E. J . 1971. Characteristics, occurrence and origins of basalts of the oceans. Abst. in Phil. Trans. Roy. Soc. Lond. A. 268: 493. Engel, A. E. J., C. G. Engel, and R. G. Havens. 1965. Chemical charac-teristics of oceanic basalts and the upper mantle. Bull. Geol. Soc. Am. 76: 719-734. Engel, A. E. J., and C. G. Engel. 1970. Mafic and ultramafic rocks in The Sea, A. E. Maxwell, Ed., Part I, 465-520, Wiley-Interscience, New York. Engel, C. G., and A. E. J. Engel. 1963. Basalts dredged from the north-eastern Pacific Ocean. Science 140: 1321-1324. Engel, C. G., and R. L. Fisher. 1969. Lherzolite, anorthosite, gabbro and basalt dredged from the Mid-Indian Ocean Ridge. Science 166: 1136-1141. Ernst, W. G. 1970. Tectonic contact between the Franciscan melange and the Great Valley sequence--crustal expression of a late Mesozoic Benioff Zone. J. Geophys. Res. 78.: 886-902. Ewing, J., M. Ewing, T. Aitken, and W. J . Ludwig. 1968. North Pacific sediment layers measured by seismic profiling in The Crust and Upper Mantle of the Pacific Area, L. Knopoff, C. L. Drake, and J. H. Pembroke, Ed., Monograph 12_, Am. Geophys. Union 147-173. Flanagan, F. S. 1967. U. S. Geological Survey silicate rock standards. Geochim. Cosmochim. Acta 31: 289-308. 214 Fleischer, R. L., and P. B. Price. 1964. Glass dating by fission fragment tracks. J. Geophys. Res. 69: 331-339. 1965. Techniques for geological dating of minerals by chemical etching of fission fragment tracks. Geochim. Cosmochim. Acta 29_: 1705-1714. Fleischer, R. L., J . R. M. Viertl, P. B. Price, and F. Aumento. 1968. Mid-Atlantic Ridge, age and spreading rates. Science 161: 1339-1342. Fletcher, W. K. 1970. Analysis of rocks and minerals by atomic absorption spectrophotometry. Unpublished personal communication. Fowler, G. A., and L. D. Kulm. 1970. Foraminiferal and sedimentological evidence for uplift of the deep-sea floor, Gorda Rise, Northeastern Pacific. J . Mar. Res. 28^ : 321-329. Frey, F. A., M. A. Haskin, J . Poetz, and L. A. Haskin. 1968. Rare earth abundances in some basic rocks. J. Geophys. Res. 73: 6085-6098. Fuller, R. E. 1932. Concerning basaltic glass. Am. Min. 17_: 104-107. Gast, P. W. 1970. Dispersed elements in oceanic volcanic rocks. Phys. Earth Planet. Int. 3_: 246. Green, D. H. 1971. Composition of basaltic magmas as indicators of conditions of origin: application to oceanic volcanism. Phil. Trans. Roy. Soc. Lond. A. 268: 707-725. Green, D. H., and A. E. Ringwood. 1967. The genesis of basalt magmas. Contrib. Mineral. Petrol. 15_: 103-190. Griggs, G. B., and L. D. Kulm. 1970. Sedimentation in Cascadia deep-sea channel. Bull. Geol. Soc. Am. 81_: 1361-1384. Grow, J . A. 1971. Rock and sediment samples from the north slope of the central Aleutian Trench. Geol. Soc. Am., Abstracts with Programs for 1971, 3: 129. Hamilton, E. L. 1967. Marine geology of abyssal plains in the Gulf of Alaska. J. Geophys. Res. 72_: 4189-4213. Hamilton, E. L., and H. W. Menard. 1968. Undistorted turbidites on the Juan de Fuca Ridge (abstract). Trans. Amer. Geophys. Union 49: 208. Hamilton, W. 1969. Mesozoic California and the underflow of Pacific Mantle. Bull. Geol. Soc. Am. 8£: 2409-2430. Hayes, D. E., and M. Ewing. 1970. Pacific boundary structure in The Sea, A. E. Maxwell, Ed., Part II, 4_: 29-72. Wiley-Interscience, New York. 215 Hayes, J. D., and D. Ninkovich. 1970. North Pacific deep-sea ash chronology and age of present Aleutian underthrusting in Geologic Investigations of the North Pacific, J. D. Hays, Ed. Mem. Geol. Soc. Am. 126: 263-290. . Hedge, C. E., and Z. E. Peterman. 1970. The strontium isotopic composition of basalts from the Gorda and Juan de Fuca Ridges, northeastern Pacific Ocean. Contrib. Mineral. Petrol. 27_: 114-120. Holmes, M. L., R. von Huene, and D. A. McManus. 1972. Seismic reflection evidence supporting underthrusting beneath the Aleutian Arc near Amchitka Island. J . Geophys. Res. 7_7: 959-964. Hood, P. 1965. Geophysical reconnaissance of Hudson Bay. I. Sea-magnetometer survey. Can. Geol. Surv. Paper 65-32: 1-11. Hurley, R. J. 1960. The geomorphology of abyssal plains in the northeast Pacific Ocean. Scripps Institution of Oceanography Tech. Rept. (S10 Reference 60-7), 105 pp. Hyndman, R. D., and D. S. Rankin. 1972. The Mid-Atlantic Ridge near 45°N. XVIII. Heat flow measurements. Can. J. Earth Sci. j): 664-670. Irving, E., W. A. Robertson, and F. Aumento. 1970. The Mid-Atlantic Ridge near 45°N. VI. Remanent intensity, susceptibility, and iron content of dredged samples. Can. J. Earth Sci._7: 226-238. Isacks, B., J. Oliver, and L. R. Sykes. 1968. Seismology and the new global tectonics. J . Geophys. Res. 73_: 5855-5900. Kay, R., M. J. Hubbard, and P. Gast. 1970. Chemical characteristics and origins of oceanic ridge volcano rocks. J . Geophys. Res. 75: 1585-1613. King, B. C. 1970. Vulcanicity and r i f t tectonics in East Africa in African magmatism and tectonics, T. N. Clifford and I. G. Gass, Ed., p. 263-283. Hafner Publ. Co., Darien, Conn. Korgen, B. J., G. Bodvarsson, and R. S. Mesecar. 1971. Heat flow through the floor of Cascadia Basin. J. Geophys. Res. 76_: 4758-4774. Kuno, H. 1950. Petrology of Hakone Volcano. Bull. Geol. Soc. Am. .61_: 957-1020. 1959. Origin of Cenozoic petrographic provinces of Japan and surrounding areas. Bull. Volcanol. 2£: 37-76. 1966. Lateral variations of basalt magma type across continental margins and islands. Bull. Volcanol. 29: 195-222. 1968. Differentiation of basalt magmas in Basalts, H. H. Hess and A. Poldervaart, Ed., 2: 623-688. Interscience Publishers, New York. 216 Langmyhr, J. F., and P. E. Paus. 1968. Analyses of silicate rocks by atomic absorption. Analytica Chim. Acta 43: 397-408. Larson, Roger L. 1971. Near-bottom geologic studies of the East Pacific Rise crest. Bull. Geol. Soc. Am. 82/. 823-842. LePichon, X. 1968. Sea-floor spreading and continental drift. J . Geophys. Res. 73: 3661-3698. Lister, C. R. B. 1970a. Heat flow west of Juan de Fuca Ridge. J . Geophys. Res. 75-14: 2648-2654. 1970b. Heat flow and sea floor spreading off the coast in Symposium on tectonism of the Pacific northwest, Ann D. Moen, Ed. EOS, Trans. Amer. Geophys. Union 52/ 628-630. Macdonald, G. A., and T. Katsura. 1964. Chemical composition of Hawaiian lavas. J. Petrol. 5/ 82-133. Mammerickx, J., and I. L. Taylor. 1971. Bathymetry of the PIONEER survey area, north of 45°N latitude, Geological Data Center, Scripps Institution of Oceanography, Special Chart No. 1. Marshall, M., and A. Cox. 1970a. Magnetism of pillow basalts and their petrology. Bull. Geol. Soc. Am. 82/ 537-552. 1970b. Weathering of deep sea basalts. Abst. in EOS (Trans. Amer. Geophys. Union) National Fall Mtg., 745. Mason, R. G., and A. D. Raff. 1961. A magnetic survey off the west coast of North America, 32°N to 42°N. Bull. Geol. Soc. Am. 72: 1259-1265. Matthews, D. H. 1971. Altered basalts from Swallow Bank, an abyssal h i l l in the Northeast Atlantic, and from a nearby seamount. Phil. Trans. Roy. Soc. Lond. A. 268: 551-571. Matthews, D. J . 1939. Tables for the velocity of sound in pure water and sea water. Hydrographic Dept., Admiralty, London. McEvilly, T. V. 1968. Sea floor mechanics north of Cape Mendocino, California. Nature 220: 901-903. McKenzie, D., and B. Julian. 1971. Puget Sound, Washington, Earthquake and the mantle structure beneath the northwestern United States. Bull. Geol. Soc. Am. 82: 3519-3524. McManus, D. A. 1964. Major bathymetric features near the coast of Oregon, Washington and Vancouver Island. Northwest Sci. 38/ 65-82. 1965. Blanco fracture zone, Northeast Pacific Ocean. Marine Geol. 3/ 429-455. 1967. Physiography of Cobb and Gorda rises, northeast Pacific Ocean. Bull. Geol. Soc. Am. 78: 527-546. 217 McManus, D. A., and OTHERS. 1970. Initial reports of the deep sea drilling project, 5_, U. S. Government Printing Office, Washington, D. C., 827 pp. McManus, D. A., M. L. Holmes, B. Carson, and S. M. Barr. 1972. Late Quaternary tectonics, northern end of Juan de Fuca Ridge (northeast Pacific). Marine Geol. 12: 141-164. Medlin, J . H., N. H. Suhr, and J . B. Bodkin. 1969. Atomic absorption analysis of silicates employing LiB02 fusion. Atomic Absorption Newsletter «S: 25-29. Melson, W. G., G. Thompson, and Tj. H. van Andel. 1968. Volcanism and metamorphism in the Mid-Atlantic Ridge, 22°N latitude. J. Geophys. Res. 73: 5925-5941. Melson, W. G., and G. Thompson. 1971. Petrology of a transform fault zone and adjacent ridge segments. Phil. Trans. Roy. Soc. Lond. A. 268: 423-441. Menard, H. W. 1960. The East Pacific Rise. Science 132 : 1737-1746. 1964. Marine Geology of the Pacific. McGraw-Hill, New York, 271 pp. 1967. Sea-floor spreading, topography, and the second layer. Science 157: 923-924. Menard, H. W., and R. S. Dietz. 1952. Submarine geology of the Gulf of Alaska. Bull. Geol. Soc. Am. 62: 239-253. Menard, H. W., and J. Mammerickx. 1967. Abyssal h i l l s , magnetic anomalies and the East Pacific Rise. Earth Planet. Sci. Lett. 2: 465-472. Menard, H. W., and T. Atwater. 1968. Changes in direction of sea floor spreading. Nature 219: 463-467. 1969. Origin of fracture zone topography. Nature 222: 1037-1040. Miyashiro, A., F. Shido, and M. Ewing. 1969. Diversity and origin of abyssal tholeiite from the Mid-Atlantic Ridge near 24° and 30°N latitude. Contrib. Mineral. Petrol. 23: 38-52. 1970. Crystallization and differen-tiation in abyssal tholeiites and gabbros from mid-oceanic ridges. Earth Planet. Sci. Lett. 7: 361-365. Moore, D. G. 1968. Reflection profiling studies of the California continental borderland: Structure and Quaternary turbidite basins. Geol. Soc. Am. Spec. Paper 107: 142. 218 Moore, G. W. 1970. Sea-floor spreading at the junction between Gorda Rise and Mendocino Ridge. Bull. Geol. Soc. Am. 81: 2817-2824. Moore, J . G. 1966. Rate of palagonitization of submarine basalt adjacent to Hawaii. U. S. Geol. Surv. Prof. Paper 550-D: 163-171. Morgan, W. J. 1968. Rises, trenches, great faults and crustal blocks. J. Geophys. Res. 73: 1959-1982. Muffler, L. J. P., J. M. Short, T. E. C. Keith, and V. C. Smith. 1969. Chemistry of fresh and altered basaltic glass from the Upper Triassic Hound Island volcanics, southeast Alaska. Am. J . Sci. 267: 196-209. Muir, I. D., and C. E. Tilley. 1964. Basalts from the northern part of the r i f t zone of the Mid-Atlantic Ridge. J. Petrol. 5_: 409-434. 1966. Basalts from the northern part of the Mid-Atlantic Ridge. II. The Atlantis collection near 30°N. J. Petrol. 7_: 193-201. Nayudu, Y. R. 1962. "A new hypothesis for the origin of guyots and seamount terraces," in Crust of the Pacific Basin, Geophys. Monograph No. 6, Nat. Res. Council, 171-180. Nayudu, Y. R., C. Chastain, and J. L. Woodruff. 1967. Morphology and petrology of two seamounts on Eickelberg Ridge (abstract). Program 63rd Annual Meeting, Geol. Soc. Am., Cordilleran Section, p. 53. Nettleton, L. L. 1962. Gravity and magnetics for geologists and seismologists. Am. Assoc. Petroleum Geol. 46: 1815-1838. Nicholls, G. D. 1965. Basalts from deep-ocean floor. Mineral. Mag. 34: 373-388. Nicholls, G. D., A. J. Nalwalk, and E. E. Hays. 1964. The nature and composition of rock samples dredged from the Mid-Atlantic Ridge between 22°N. and 52°N. Marine Geol. 1_: 333-343. Paster, T. P. 1971. Petrologic variation within submarine basalt pillows of the south Pacific Ocean in Antarctic Oceanology I, Amer. Geophys. Union Antarctic Res. Ser. 15/ 283-308. Pavoni, N. 1966. Tectonic interpretation of the magnetic anomalies southwest of Vancouver Island. Pure Appl. Geophys. 63_: 172-177. Peacock, M. A. 1928. The petrology of Iceland. Part 1. The basic tuffs. Trans. Roy. Soc. Edin. 5_5_: 53-76. Peacock, M. A., and R. E. Fuller. 1928. Chlorophaeite, sideromelane, and palagonite from the Columbia River Plateau. Am. Min. 13_: 360-383. 219 Peter, G., and R. Lattimore. 1969. Magnetic structure of the Juan de Fuca Gorda Ridge area. J . Geophys. Res. 74: 586-593. Peter, G., and 0. E. DeWald. 1971. Deformation of the sea floor off the northwest coast of the United States. Nature 232: 97-98. Peters, L. V. 1949. The direct approach to magnetic interpretation and its practical applications. Geophysics 14_: 290-320. Raff, A. D., and R. G. Mason. 1961. Magnetic survey off the west coast of North America, 40°N latitude to 50°N latitude. Bull. Geol. Soc. Am. 72: 1267-1270. Ringwood, A. E. 1969. Composition of the crust and upper mantle in The Earth's Crust and Upper Mantle, P. J. Hart, Ed., Amer. Geophys. Union Monograph 13_: 1-17. Schilling, J. G. 1971. Sea-floor evolution: rare-earth evidence. Phil. Trans. Roy. Soc. Lond. A. 268: 663-706. Schneider, E. D., and P. Vogt. 1968. Discontinuities in the history of sea-floor spreading. Nature 217: 1212-1222. Scholl, D. W., M. N. Christensen, R. von Huene, and M. S. Marlow. 1970. Peru-Chile trench sediments and sea-floor spreading. Bull. Geol. Soc. Am. 81_: 1339-1360. Shor, G. G., P. Dehlinger, H. K. Kirk, and W. S. Franch. 1968. Seismic refraction studies off Oregon and Northern California. J . Geophys. Res. 73: 2175-2194. Shouldice, D. H. 1971. Geology of the western Canadian continental shelf. Bull. Can. Petroleum Geol. 1£: 405-436. Silver, E. A. 1969. Late Cenozoic underthrusting of the continental margin off northernmost California. Science 166: 1265-1266. 1971a. Transitional tectonics and late Cenozoic structure of the continental margin off northernmost California. Bull. Geo. Soc. Am. 8_2: 1-22. 1971b. Tectonics of the Mendoceno triple junction. Bull. Geol. Soc. Am. 82: 2965-2978. 1971c. Small plate tectonics in the northeastern Pacific. Bull. Geol. Soc. Am. 82: 3491-3496. Silver, E. A., and W. D. Barnard. 1971. Pleistocene tectonics of the Washington continental slope. Geol. Soc. Am., Abstracts with Programs for 1971, 3_: 193. Silvestri, S. 1963. Proposal for a genetic classification of hyaloclastites. Bull. Volcanol. 25: 315-321. 220 Slemmons, D. B. 1962. Determination of volcanic and plutonic plagioclases using a three- or four-axis universal stage. Geol. Soc. Am. Spec. Paper 69, 64 pp. Smith, J . V., and R. C. Stenstrom. 1965. Chemical analysis of olivines by the electron microprobe. Mineral Mag. 34_ (Tilley Volume) 436-459. Smith, R. E. 1967. Segregation vesicles in basaltic lava. Am. J. Sci. 265: 696-713. Suhr, N. H., and C. 0. Ingamells. 1966. Solution technique for analysis of silicates. Anal. Chem. 38_: 731-734. Talwani, M., X. LePichon, and J. R. Heirtzler. 1965. East Pacific Rise: The magnetic pattern and the fracture zones. Science 150: 1109-1115. Tatsumoto, M., C. E. Hedge, and A. E. J. Engel. 1965. Potassium, rubidium, strontium, thorium, uranium, and the ratio of strontium-87 to strontium-86 in oceanic tholeiitic basalt. Science 150: 886-888. Ti f f i n , D. L. 1970. Continuous seismic reflection profiling in the Strait of Georgia, British Columbia. Ph.D. dissertation, University of British Columbia, 177 pp. Tif f i n , D. L., B. E. B. Cameron, and J. W. Murray. 1972. Tectonics and depositional history of the continental margin off Vancouver Island, British Columbia. Can. J . Earth Sci. £: 280-296. Tobin, D. G., and L. R. Sykes. 1968. Seismicity and tectonics of the northeast Pacific Ocean. J. Geophys. Res. 72: 3821-3846. Udintsev, G. B., L. V. Dmitriev, and A. P. Vinogradov. 1971. The tectonics of the Mid-Indian Ocean Ridge and the petrography of the solid rocks of its r i f t zones. Phil. Trans. Roy. Soc. Lond. A. 268: 653-659. United States Coast and Geodetic Survey. 1970. Seismicity of the United States, NEIC 3012. Vallance, T. G. 1965. On the chemistry of pillow lavas and the origin of spilites. Mineral Mag. 34: 471-481. van Andel, Tj. H. 1969. Recent uplift of the Mid-Atlantic Ridge south of the Vema Fracture Zone. Earth Planet. Sci. Lett. 7_: 228-230. van Andel, Tj. H., and C. 0. Bowin. 1968. Mid-Atlantic Ridge between 22° and 23°N and the tectonics of mid-ocean rises. J. Geophys. Res. 73: 1279-1298. van Andel, Tj . H., and G. R. Heath. 1970. Tectonics of the Mid-Atlantic Ridge 6° to 8°S latitude. Marine Geophys. Res. 1_: 5-36. 221 Vine, F. J. 1966. Spreading of the ocean floor: New evidence. Science 154: 1405-1415. 1968. Magnetic anomalies associated with mid-ocean ridges in The history of the Earth's crust, R. A. Phinney, ed., Princeton Univ. Press, p. 73-89. Vine, F. J., and J . T. Wilson. 1965. Magnetic anomalies over a young oceanic ridge off Vancouver Island. Science 150: 485-489. von Huene, R., L. D. Kulm, and OTHERS. 1971. Deep sea drilling project Leg 18. Geotimes. October. 12-15. Watkins, N. D., and A. Richardson. 1968. Comments on the relationship between magnetic anomalies, crustal spreading and continental drift. Earth Planet. Sci. Lett. 4_: 257-264. Wilcox, R. E. 1965. Volcanic-ash chronology in The Quaternary of the United States, H. E. Wright, Jr., and D. G. Frey, ed. A review volume for the 7th Congress of the International Association for Quaternary Research, Princeton, N. Y., 807-816. Wilson, J . T. 1965. Transform faults, ocean ridges, and magnetic anomalies southwest of Vancouver Island. Science 150: 482-485. Yoder, H. S., and C. E. Tilley. 1962. Origin of basaltic magmas: An experimental study of natural and synthetic rock systems. J. Petrol. 3: 342-532. 222 APPENDIX A TABLE XII. Seismic reflection profiles reproduced in Appendix A and Figure 23. Locations of profiles are shown in Figures 51 and 52. Key to symbols used on profile interpretations is in Figure 53 Profile IOUBC Profile Plate Feature A-A' PZ69-9 I Juan de Fuca Ridge B-B' 70-16-5 II Juan de Fuca Ridge C-C 70-16-6 III Juan de Fuca Ridge D-D* 70-16-4 IV Juan de Fuca Ridge E-E' PZ69-8 V Juan de Fuca Ridge F-F' 70-16-3a VI Juan de Fuca Ridge (Middle Ridge) G-C 70-16-3b VI Juan de Fuca Ridge (Middle Ridge) G-H 70-16-3C VII Juan de Fuca Ridge (Middle Valley) H-H' 70-16-3d VII Sovanco Fracture Zone I-I1 PZ69-lb VIII Sovanco Fracture Zone to Heck Seamount J-J* 70-16-8a IX Sovanco Fracture Zone J-K' 70-16-8b X Sovanco Fracture Zone K-K1 70-16-8C XI Sovanco Fracture Zone L-L1 71-23-2 Fig. 23a Sovanco Fracture Zone to Heck Seamount Chain M-M' 71-15-35 Fig. 23b Sovanco Fracture Zone to Heck Seamount Chain N-N' 70-16-7 XII South from the eastern end of Heck Seamount Chain 0-0' 71-15-lB XIII Lower continental slope P-P1 71-15-3B XIII Lower continental slope Q-Q- PZ69-7 XIV Continental slope R-R1 PZ69-la XV Continental slope S-S" 70-16-2 XVI Continental slope T-T' 71-15-4B XVII Lower continental slope U-U1 PZ69-10 XVIII Lower continental slope V-V 70-16-1 XIX Continental slope 223 FIGURE 51. Locations of seismic reflection profiles in the area of the northern end of Juan de Fuca Ridge reproduced in Plates I to XII. Bathy-metry is from Figure 15. 224 FIGURE 52. Locations of seismic reflection profiles over the continental slope reproduced in Plates XIII to XIX. Bathymetry is from Figure 15. 225 VERTICAL SCALES ON PROFILES ARE IK SECONDS OF 2-WAY TRAVEL TIME IGNEOUS BASEMENT (dashed where uncertain) GOOD REFLECTOR IN SEDIMENTS MODERATE REFLECTOR IN SEDIMENTS POOR REFLECTOR IN SEDIMENTS A SEDIMENTARY UNIT A B SEDIMENTARY UNIT B MAJOR FAULT (displacement of Igneous basement i s generally obvious) MINOR FAULT (displacement or disruption of sedimentary sequence Is obvious; displacement of basement not generally obvious) ? UNCERTAIN DREDGE STATION REPRESENTATIVE NOMOGRAPHS to i l l u s t r a t e the ve r t i c a l exaggerations of the seismic reflection profiles. Vertical exaggerations and nomographs calculated using a velocity of sound in water of 1.5 km/sec. V.E. x 6.5 V.E. x 10.5 V.E. x 13.5 FIGURE 53» Key to seismic refle c t i o n p r o f i l e s , Plates I-XII and Figure 23. PLATE II "Sis i «1 * 8 * I ?*0 PLATE III PLATE IV PLATE V 231 CM (O + S V W W V O 1 S Q N O 0 3 s 232 s Q N O 0 3 s I—I—I / I I I o o o o o CSI CM svwwv'o • — | J C » . " J . L co m E if — I CO X — J Ml r UIWSflL mm Mm. J s a N O 0 3 s 234 PLATE IX 2 3 5 PLATE X 236 PLATE XI 237 > H a x i—i i — i h-1 8£3 239 4i. 242 PLATE XVII 243 PLATE XVIII PLATE XIX 245 APPENDIX B PART 1 PETROGRAPHIC DESCRIPTIONS OF BASALTIC ROCKS The petrographic descriptions are presented under their separate dredge station headings. The stations are arranged in numerical order, and locality headings are also given. Locations of the dredge stations are shown on Figure 28. Details of the hauls are given in Table XIII (p. 262). Numerous references are made to photographs in Figures 30 to 36, Chapter IV. Juan de Fuca Ridge Station 1 (Middle Ridge) Nearly a l l the rocks from Station 1 on the western flank of Middle Ridge near its northern end are angular fragments of glomeroporphyritic basalt. This basalt is massive and only sparsely vesicular. Most of the samples appear weathered, and have a brown or greyish-brown oxide coating with a maximum thickness of 1 mm. The rocks split along irregular fractures, whose surfaces are stained with yellow-brown iron hydroxides and speckled with dendritic, black manganese oxide growths. In some more weathered samples, the fracture surface is coated with yellow-brown to grey-brown clay. On a cut surface, large plagioclase crystals are observed scattered through a microcrystalline grey matrix (Fig. 30a). These large crystals are thought to be xenocrysts (see p. 115). In the interiors of large specimens the plagioclase crystals appear greyish-white and fresh; however, near the 246 margins and in small specimens, the crystals are yellowish. The proportion of xenocrysts (which generally occur in clusters) varies from 3 to 20% but is commonly about 15 to 20%. A few specimens have black glassy margins of varying thicknesses to 1 cm. The glass has concentric fractures, in places f i l l e d by yellow palagonite, and the outer part of the glass has been altered to palagonite. Xenocrysts also occur in the glass. In thin section, the plagioclase xenocrysts are generally subhedral to rounded, and some have ragged edges. They generally contain patches and stringers of tachylite, strung out along the 010 cleavage and twin plane (Fig. 30c). Similarly occurring glass inclusions (Melson and Thompson, 1971), as well as fluid inclusions (Aumento, 1968) and opaque minerals (Bertrand, 1972), have been described in other porphyritic submarine basalts. The plagioclase xenocrysts show twinning, commonly of the carlsbad and albite twin laws. Some are strongly zoned, but others show only narrow zones near the outer rims and some show no zoning except an outer rim of more sodic com-position. The inner zones of xenocrysts average about An85, but values as high as An90 were measured. The outermost zones of xenocrysts are about An60. Smaller xenocrysts (phenocrysts?) are about An75-80. It should be pointed out that i t is very difficult to determine the compositions of narrow zones and small laths accurately using the universal stage techniques, so that errors in outer zone and groundmass lath determinations are probably as great as ±10% in some cases. Pseudomorphs of olivine crystals occur rarely in the basalt, the original olivine, of which a tiny relict sometimes survives, having been replaced by greenish-brown alteration products (serpintine?) in places with 247 associated carbonate. They f i l l or line the generally euhedral to subhedral olivine form (Fig. 30b). The groundmass is tachylite. It is brown or greyish-brown in plane light> but under crossed nicols i t is slightly birefringent, probably due to the presence of cryptocrystalline clinopyroxene. In patches i t is altered to greenish-brown alteration products (serpentine?) similar to those replacing olivine. Such alteration products may constitute over 5% of the basalt. Set in this obscure mesostasis are a multitude of small plagioclase micro-lites, in places skeletal and f i l l e d with tachylite (Fig. 30b, c). These microlites are rarely twinned. The larger laths are about An70-75, but most are An60-70, with a narrow outer zone probably between An50 and An60. The smallest microlites are probably also in this latter compositional range. The interior texture of this basalt is glomeroporphyritic hyalopilitic to intersertal. Station 2 (Middle Ridge) This station is about 1.5 km north of Station 1 on the western flank of Middle Ridge, and about 300 m deeper. About 135 kg of basaltic rocks were collected, none of which are porphyritic basalts like those from Station 1. The samples are divided into two groups: i . Holocrystalline basalt with intergranular texture i i . Fine-grained basalt with glass margins i . Holocrystalline basalt with intergranular texture About half the samples obtained at Station 2 are massive, smooth-surfaced but angular pieces, with stain (green-brown-yellow) on their outer surfaces and along broken surfaces (Fig. 30d). No oxide coating has developed. These blocky rocks are non-glassy. They have well-developed alteration rims 248 of variable thicknesses which look dark and yellowish-brown compared to the fresh medium-grey interiors. Rare vesicles are round and large (0.5 to 1 mm in diameter). They are usually lined by bluish-grey or green material, but where they occur in the alteration rims they may be fil l e d with green and yellow radiating material. In thin section the massive basalt has an intergranular texture (Fig. 30e) with small plagioclase laths (An50-57) intergrown with fine-grained clinopyroxene. A few plagioclase microphenocrysts (An60) are present. The basalt contains about 3% small opaque minerals. There is also minor brown interstitial material, birefringent under crossed nicols, which may be an alteration product of pyroxene, or of a residual cryptocrystalline groundmass. i i . Fine-grained basalt with glass margins Most of the other specimens obtained at Station 2 are of this type. Many specimens are angular and irregular, but others have the characteristic "broken pillow fragment" shape, with one curved, glassy surface. Many are partly encased in a rind of palagonite breccia, giving a very rubbly surface texture (Fig. 30d). This breccia consists of irregular chips of glass, and, less commonly, very fine-grained basalt fragments, set in a yellowish matrix of palagonite and its alteration products. The specimens which are not coated by the breccia have patchy dark oxide coatings up to 1 cm in thickness. The fine-grained basalts tend to break along stained fractures which have dendritic manganese oxide growths. Some specimens have wide altera-tion rims, in places consisting of more than one concentric band, which are darker than the medium-grey interiors. Most contain many (up to 5%) small, round or irregularly shaped vesicles, some of which are f i l l e d or lined by green or yellow chlorophaeite. 249 In thin section the specimens with glass selvages show the typical transition from pale brown, isotropic sideromelane through a variolitic zone to a more crystalline interior where the texture is hyalopilitic, intersertal, or even, in patches, intergranular. In the interiors of the fragments, the mesostasis consists of sheaves of cryptocrystalline clinopyroxene and feldspar with scattered small.opaque minerals in the interstices (Fig. 30f). Plagioclase also occurs as laths and microlites of varying sizes, in many cases skeletal with tachylite cores. The larger plagioclase laths, most of which have narrow, more sodic rims gave compositions in the range An70-An60, whereas smaller laths were An45-An60. Scattered throughout are small, skeletal olivine crystals, usually at least partly replaced by a peseudo-morphic greenish-brown alteration product--serpentine?--which is also common in the interstices of the basalt. Some specimens contain an unusually high percentage of opaque minerals (Fig. 30f) about 10-15%. This can be correlated with high iron and titania contents in the chemical analyses of these rocks (see p. 127). Vesicles are of irregular shapes, and empty in the interior regions of the specimens. Near the rims they are generally f i l l e d with yellow or reddish-yellow or greenish-yellow material (chlorophaeite). Some specimens of breccia-coated and glassy basalt were recovered which are less massive than the other rocks described. They contain large (5 mm or more in maximum dimension), very irregular vesicles, as well as scattered small, round vesicles. Station 3 (West Ridge) About 100 kg of glomeroporphyritic olivine basalt were dredged from the western flank of West Ridge. All are obvious pillow fragments, with the 250 characteristic curved glassy surfaces and radial jointing. The curved surfaces have yellowish-brown palagonite rinds. Under the outer palagonitized rind is a thick (1-2 cm) layer of fresh-looking black glass. The basalt consists of 5-10% clusters of olivine crystals in a very fine-grained dark matrix (Fig. 31a). These crystals are interpreted to be xenocrysts (see p. 115-117). The glass selvage also contains xenocrysts. In thin section, the interior texture is seen to be glomeroporphy-r i t i c hyalopilitic (Fig. 31b), with dark, weakly birefringent tachylite forming a mesostasis in which sit plagioclase laths and microlites of various sizes, and euhedral to anhedral olivine crystals. In the glass rim, the glass is sideromelane, and the typical transition through the variolitic zone to the hyalopilitic interior is observed. The plagioclase laths are zoned, but the compositional range is not large (An60-70). The olivine com-position (determined by an X-ray diffractometer method--Smith and Stenstrom, 1965), is about Fo92. These basalts are very fresh, and no alteration products were observed in thin section. Station 4 (Endeavour Seamount) The only sample recovered from the southwestern flank of Endeavour Seamount is a bulbous, sinuous pillow. It was delivered on deck in two pieces which f i t together at a narrow neck (Fig. 31c). Two other round to elliptical broken surfaces with radial fractures are present. One has a thin coating of brown oxides and does not look like a new break. However, the other has only stain like that which occurs on the two surfaces which f i t together, and was perhaps the surface of attachment, broken by the dredge. Such pillows must form strange contorted outcrops. Underwater movies in the region have shown similar forms on the sea floor (C. R. B. Lister, pers. comm., 1972). 251 An outer yellowish-brown palagonite layer up to 5 mm in thickness spalled off readily, exposing a fresh black glass surface with polygonal fracture patterns produced by shrinkage during cooling. This surface is dotted with white plagioclase crystals up to 1 cm across (Fig. 31d, c). These are interpreted to be xenocrysts (see p. 115-117). A few smaller green pheno-crysts of olivine are associated with the plagioclase, and, like the plagioclase, are unaltered. Despite the large size of the pillow, the basalt is cryptocrystalline throughout. In thin section the basalt is observed to be variolitic. In plane light the varioles are dark reddish-brown or greyish in colour, but are birefringent under crossed nicols. A few microlites of plagioclase and small olivine and clinopyroxene crystals are scattered through the variolitic mesostasis. The large plagioclase xenocrysts are subhedral to anhedral in outline and commonly show twinning (generally carlsbad) and zoning. The inner zones of large xenocrysts are An85-90, while smaller laths are about An82. Although none could be accurately determined, outer zones are estimated to be about An60. Sovanco Ridge Station 5 Approximately 150 kg of basalt were recovered from the base of the fault scarp at the eastern end of Sovanco Ridge (see seismic Profile G-H--Plate VII). These are divided into three main groups: i . Angular, smooth-surfaced blocks of holocrystalline basalt i i . Fine-grained and glassy porphyritic basalt i i i . Fine-grained and glassy non-porphyritic basalt 252 i . Angular, smooth-surfaced blocks of holocrystalline basalt Over one-third of the specimens belong to this group. The fresh surfaces are medium to dark grey, except near the margins where a darker, \ yellow-brown stained alteration rim is present (Fig. 32a). The group can be subdivided, one subgroup being more vesicular and more weathered-looking, with wider alteration rims. The vesicles are small, and in some cases fi l l e d or lined with blue-green material. In thin section, the texture is intergranular (Fig. 32b), with small crystals of clinpyroxene forming the matrix for larger plagioclase laths. Scattered opaques and patches of brown alteration products in the interstices form a small portion of the basalt. No phenocrysts were observed. The plagioclase compositions determined ranged from An62 for small laths to An73 for the inner zone of one of the larger laths. The specimens of the other subgroup also have intergranular textures (Fig. 32c) but the grain sizes, particularly of the clinopyroxenes, are coarser. Olivine has generally been pseudomorphed by brown alteration products which also f i l l interstices. About 3% scattered opaques occur. These specimens contain a few skeletal phenocrysts of plagioclase which contain stringers of tachylite. The dark alteration rims are the result of chlorophaeite and iron oxides and hydroxides f i l l i n g the vesicles and interstices (see p. 149-154). i i . Fine-grained and glassy porphyritic basalts About one-third of the specimens from Station 5 belong to this group. The glass selvages are black and lustrous where fresh, and over 1 cm in thick-ness. However, the outermost glass has been altered to palagonite (Fig. 32d). Alteration has also penetrated along fractures into the interiors of many specimens (Fig. 32d). 253 In thin section the typical transition from clear, pale brown sideromelane through a variolitic zone to a hyalopilitic interior is observed. Plagioclase xenocrysts, which constitute between 10 and 20% of the basalt, are commonly zoned, contain stringers of tachylite, and show strong effects of resorption (Fig. 32e, f) . The composition of the inner zones of the xenocrysts is about An85, while outer zone measurements varied from An62 to An68. Most xenocrysts have a very narrow outermost, more sodic rim. The groundmass consists of birefringent tachylite with crypto-crystalline pyroxenes and a network of plagioclase laths. The plagioclase laths are in many cases hollow, with cores of tachylite. Their compositions lie in the range An60-70. Like the porphyritic basalts of Station 1, these basalts show strong effects of alteration, with greenish-brown, serpentine-like alteration products pseudomorphic after olivine and f i l l i n g interstices. i i i . Fine-grained and glassy non-porphyritic basalt These specimens have well-developed glass rinds, generally curved, with the outer glass palagonitized. Palagonite has also developed within concentric fractures in the glass selvage. The basalt contains many small, round vesicles. In thin section, the samples show the typical transition from pale brown sideromelane through a variolitic zone to a hyalopilitic interior (Fig. 33a, b, c). In the sideromelane, vesicles are in many cases rimmed by incipient variolitic growths (Fig. 33a). They are also often lined with yellow fibro-palagonite (Fig. 33b), or tachylite and palagonite (segregation vesicles--Smith, 1967). Variolite development also occurs around plagioclase microlites (Fig. 33b). 254 In the hyalopilitic interiors, the plagioclase laths tend to show a rough alignment (Fig. 33c). A few scattered olivine phenocrysts occur, usually fresh but some are partly altered to green-brown serpentine (?) (Fig. 33c). Station 6 Station 6 is on the southern flank of one of the peaks forming the crest of Sovanco Ridge (see Profile L-L'--Fig. 23a). Approximately 750 kg of basalt were recovered, but they apparently show less variation than is observed among the samples from Station 5. Nearly all the rocks recovered are pillow fragments, with one convex, glassy surface and other faces broken along fractures roughly normal to the convex surface to form wedge-shaped blocks (Fig. 34a). A few are bulbous, more complete pillow shapes with radial fractures. Many other pieces are convexo-concave slabs 3 to 8 cm in thickness. In many cases both surfaces are glassy, although the convex surface is typically rough with polygonal fractures, while the concave glass surface is smooth. One slab has a ridge on the concave surface where lava may have flowed into a crack on the under-lying surface. These platy lava fragments may be pieces of single flows over cool rocks. Or they may have formed crusts from below which lavas were periodically withdrawn, as on the surface of a former lava lake. Similar slabs recovered from the Mid-Atlantic Ridge were explained in this way by Melson and Thompson (1970). Also present are a few small, fine-grained fragments completely or partly encased in palagonite breccia as described at Stations 2 and 5 (Fig. 34b). Such fragments probably form part of the interpillow breccias. 255 A few rocks have oxide coatings up to 1 cm in thickness. One specimen has a thick (1 cm) glass margin on one side and a thick (1 cm) oxide coating on the other. This slab perhaps f e l l from its original cooling position so that original "bottom" was exposed to oxide deposition. Most of the specimens from Station 6 are very vesicular--these rocks as a group show the greatest degree of vesicularity of any rocks dredged from the study area. Most cut surfaces show at least 5% small, round or elliptical vesicles. Most are empty, but a few are lined with yellow-red iron oxides, or blue or green minerals. The most vesicular specimens contain about 10% small empty vesicles. In thin section the basalts from Station 6 reveal great textural variation, but are characterized by the presence of fresh olivine micropheno-crysts, which vary from euhedral to subhedral in shape (Fig. 34c). The outer glass is sideromelane, which becomes variolitic towards the interiors. Many varioles radiate from plagioclase microlites in their cores. In the interiors of the specimens, the texture may be intersertal or hyalopilitic (Fig. 34d, e, f ) . The groundmass plagioclase laths range from about An60 to An70. Some specimens exhibit a peculiar "oolitic" texture developed in bands (Fig. 34a). In thin section, these "oolites" are observed to be due to radiating sheaves of tiny crystals, probably plagioclase with intergrown clinopyroxene. One medium-grained basalt (specimen 6-7) was found in the assemblage from Station 6. It contains about 3% vesicles and is highly altered, with yellow-brown bands. The freshest part, in the core, is medium grey. In thin section, the texture is observed to be intergranular, with pale brown augite crystals, opaque minerals, and olivine in a framework of plagioclase laths. 256 The olivine is fresh, but anhedral (rounded) and fragmental, with brown alteration around the edges and along cracks. The highly altered bands are rich in yellowish- or greenish-brown chlorophaeite in vesicles and inter-stices, contain unusually high concentrations of opaque minerals, and have darker brown pyroxene crystals. Segregation vesicles, lined with tachylite, alteration products, and minor carbonate, are common (Smith, 1967). Heck Seamount Chain Station 7 At the eastern end of the Heck Seamount Chain the pipe dredge recovered about 50 kg of basalt, mostly very small fragments, mixed with greenish-brown foraminifer-rich mud. These basalts are very fresh with no oxide coatings. Most of the fragments are clinker in small, curled, rolled, and bulbous forms (Fig. 35a). Larger chunks tend to be jagged, scoriaceous, and cavernous. One surface, presumably the upper, is usually oxidized to reddish-brown. Many small fragments consist of black, lustrous glass through-out, and tubular vesicles are common. The exterior glass surface is generally palagonitized. Occasional plagioclase phenocrysts, and more rarely olivine pheno-crysts, are observed on the cut or broken surfaces (Fig. 35b). In thin section the groundmass is generally observed to be crypto-crystalline with a variolitic texture. Yellow-green chlorophaeite and other alteration products line and f i l l vesicles, and are common in interstices of the groundmass. The rare olivine phenocrysts have compositions of about Fo80. More common are plagioclase phenocrysts and microphenocrysts, inner zones of which are about An80 in composition. 257 In addition to the very fine-grained and glassy basalt, one angular, holocrystalline piece of basalt was recovered (Fig. 35c). It has a prominent dark alteration rim. It contains a few scattered plagioclase phenocrysts. In thin section, the texture of this basalt is observed to be intergranular, with small plagioclase, clinopyroxene, and opaque crystals (Fig. 35d). A tentative composition for the plagioclase is An50. Plagioclase phenocrysts are as calcic as An82, although smaller ones are about An70. Vesicles in the interior are empty, but those in the alteration rim are lined or f i l l e d with green chlorophaeite. One coarse-grained rock was recovered (specimen 7-5). It is vesicular and altered with a yellow-brown iron oxide stain rimming and per-meating crystals. The basalt is composed of plagioclase, clinopyroxene, olivine, and minor opaque minerals, forming an ophitic texture, with large, optically continuous pyroxene and olivine crystals enclosing smaller plagio-clase laths. Plagioclase compositions determined ranged from An55 to An80. A vesicle-rimmed nodule, about 1 cm in diameter, occurs in a small, jagged fragment of cryptocrystalline basalt (specimen 7-6). In thin section, the nodule is seen to have an intergranular texture with a framework of large plagioclase laths f i l l e d by clinopyroxene, minor olivine, and opaques. The latter f i l l angular interstitial areas. The host rock is cryptocrystalline, but not variolitic. Both the nodule and the coarse-grained fragment are probably cognate xenoliths. Station 8 (Heck Seamount) From the north slope of Heck Seamount the pipe dredge was recovered nearly f u l l with a heterogeneous mixture of basalt, siliceous sponges, and mud. 258 The sponges were dead, and impregnated with black manganese oxide deposits which coat the clear glass spicules of the sponge structure, and which can be chipped off with a needle. In some specimens the coating is so thick that the sponge is unrecognizable. The basalt fragments are for the most part angular, with jagged, cavernous forms. Most are glassy, but cannot be readily identified as pillow fragments (Fig. 35e). Some have ropy surfaces with yellow-brown palagonitized glass on the exterior surfaces. The inner glass is black, lustrous, and fresh. However, some specimens have accumulations of oxides, usually thin and patchy but reaching a maximum thickness of 0.2 cm. Other specimens show a swirly colour banding apparently due to alteration along tiny fractures. The basalts are microcrystalline to glassy and contain vesicles of all sizes from minute to large tubular cavities. Many small chunks consist entirely of glass, palagonitized on the outer surface, and containing con-centric fractures. In thin section, the glass is pale brown, perfectly isotropic sideromelane, cut by irregular fractures, probably as a result of contraction during cooling, like the megascopic concentric fractures (Fig. 35f). Minor crystallization has occurred in the form of birefringent variolitic growths. The more crystalline basalts have sideromelane selvages and crypto-crystalline interiors where the texture is variolitic. A few small phenocrysts of plagioclase (about An75) are scattered throughout. Station 9 (Heck Seamount) The southern slope of Heck Seamount near its summit gave a very different assortment of material from the northern slope (Station 8) or the western part of the summit (Station 10). The recovery consisted of small chips of basalt and glass, and chunks of aquagene tuff (see p. 156-159) in a 259 foraminifer-rich mud. No chips of basalt were large enough for study. However, the largest fragment of glass recovered (a slab about 2.5 cm square) was used for chemical analysis. The glass contained a few small plagioclase phenocrysts, and was palagonitized on the exterior. The aquagene tuffs apparently form much of the sea floor at Station 9, because some have sponges growing on one surface, or patches of oxides. Station 10 (Heck Seamount) Only one piece of basalt was recovered from the western part of the summit of Heck Seamount. This is a 3 kg chunk of massive basalt with a glass selvage about 1 cm in thickness at one end, and a discontinuous and narrow dark alteration rim. The specimen contains a few plagioclase phenocrysts. In thin section, the texture of this basalt is observed to be intersertal to intergranular, with small plagioclase laths (An60) in a matrix of small clinopyroxene crystals and opaque minerals (Fig. 36a). The grain size of the pyroxene varies—usually i t is cryptocrystalline and turbid-looking, but in patches i t is well-crystallized and the texture is intergranular. Other material from Station 10 consists of a rubbly mixture of sponge fragments, basaltic debris, tuffaceous material, and multi-coloured sedimentary deposits (see p. 161). Station 11 The eastern slope of the western peak of the Heck Seamount Chain gave a small recovery of irregular pieces of fine-grained basalt. Most pieces have glass rims which are palagonitized on the exterior, and thick oxide coatings up to 0.6 cm in thickness (Fig. 36b). Vesicles, large, tubular, and inter-260 connected, are numerous in these specimens. On broken surfaces, swirly, coloured patterns of alteration, apparently related to tiny fractures, are observed on several specimens. Most of these basalts are cryptocrystalline with variolitic textures. The swirly banding mentioned above is due to varying variolite sizes, and the presence or absence of greenish-yellow chlorophaeite and iron oxides in the interstices and in vesicles. The freshest-looking specimen (11-1) is fine-grained but holocry-stalline, with plagioclase laths (An50-60), granular clinopyroxene crystals, and opaques forming a sub-ophitic texture, virtually the same as that of specimen 7-1 (Fig. 35d). The basalt contains about 2% irregular vesicles, empty except in the alteration rim where they are f i l l e d with green chloro-phaeite. A few small plagioclase phenocrysts (An70 with narrow, more sodic outer zones) are scattered throughout. Heckle Seamount Chain Station 12 The basalts from near the eastern end of Heckle Seamount Chain are all fragments of large, ellipsoidal pillows. All have the typical curved glassy surfaces and radial jointing. Particularly on the curved, glassy surfaces, oxide coatings are as thick as 5 mm. The glass itself is extensively altered to palagonite beneath the oxide coating. The basalts are very fine-grained, and on a cut surface i t is observed that only cores of medium-grey fresh basalt remain unaltered (Fig. 36c). In thin section, the interior texture of these basalts is inter-sertal, with patches where the pyroxene is better crystallized and the texture is intergranular (Fig. 36d). Vesicles are empty. However, in the darker 261 regions (alteration rims) the vesicles and interstices are fi l l e d with green chlorophaeite and iron oxides. These dark, altered regions may extend far into the rocks along fractures. Near glass selvages, the texture is variolitic and cryptocrystalline with some microphenocrysts (many in glomeroporphyritic clusters) of plagio-clase and clinopyroxene. The plagioclase compositions range from An70-80. Station 13 The western peak of the Heckle Seamount Chain yielded only one basalt specimen. Specimen 13-1 is an angular piece of fine-grained, non-porphyritic basalt, which is completely encased in an oxide coating (Fig. 36e). The specimen has a wide, dark alteration rim, and a fresh lighter grey core. There is a thin rim of glass at one end. The texture is similar to that of basalts from Station 12, inter-sertal with intergranular patches. Particularly towards the glass margin, the groundmass consists of varioles and sheaves of cryptocrystalline pyroxene and plagioclase. The composition of groundmass plagioclase laths is about An60, whereas a plagioclase microphenocryst is zoned from An80 to An75 with a more sodic outermost zone. As usual, in the dark rim the interstices and vesicles are f i l l e d with green chlorophaeite. TABLE X I I I . Summary description of dredge hauls in which samples of underlying bedrock were obtained DREDGE L 0 C A T I 0 N DEPTH STATION Feature N. Lat. W. Long. From To BRIEF DESCRIPTION AND REMARKS 70-1 •16- 2 Juan de Fuca Ridge: Middle Ridge, west flank 48°44' 128" 33' 2012 1865 Pipe dredge. 100 kg total. Glomeroporphyritic plagiocalse basalt, altered to varying degrees. A few chunks have glass . selvages. Groundmass cryptocrystalline. Patchy oxide coatings to 1 mm in thickness. Also small fragments of palagonite breccia, and many gravel-sized pieces of basalt in mud. 71-2 -23- 1 Juan de Fuca Ridge: Middle Ridge, west flank 48°45' 128° 34.5' 2377 2103 Chain bag dredge. 135 kg total. 70 kg of angular smooth-sided chunks of massive, holocrystalline basalt, with dark alteration rims. 65 kg of fine-grained glassy basalt, generally encased in palagonite breccia. Some pillow fragments. Typically vesicular. 71-3 -23- 3 Juan de Fuca Ridge: West Ridge, west flank near summit 48°31' 128°55' 2652 2377 Chain bag dredge. 90 kg total. Fresh glomeroporphyritic olivine basalt. Pillow fragments with 1-2 cm wide selvages of black glass, with outer palagonitized rinds. No oxide coatings. 71-4 -15- 8 Juan de Fuca Ridge: Endeavour Seamount, southwest flank 48°18.4' 129° 04.01 2377 1829 Chain bag dredge. 25 kg total. Two sinusoidal, bulbous pillow chunks which f i t together at a narrow neck. Composed of fresh porphyritic plagioclase basalt, with a crypto-crystalline groundmass. Fresh black glass selvage with outer palagonitized rind which spalls off. Basalt also contains scattered fresh olivine phenocrysts. No oxide coatings. 71-5 •15- 5 Sovanco Ridge: Base of eastern fault scarp 48°52.8' 128° 51.4' 2505 2377 Chain bag dredge. 150 kg total. Heterogeneous mixture—probably a talus deposit. 60 kg angular chunks of massive, medium-grained basalt with distinct dark alteration rims. 50 kg glassy, glomeroporphyritic plagioclase basalt with cryptocrystalline groundmass. Outer glass is palagonitized, and basalt is extensively altered. 40 kg fine-grained non-porphyritic basalt with glass selvages. No oxide coatings. TABLE XIII--Continued DREDGE L 0 C A T I 0 N DEPTH On) STATION Feature N. Lat. W. Long. From To BRIEF DESCRIPTION AND REMARKS 6 71-23-7 Sovanco Ridge: western peak, southern flank 48°55.3' 129°23.4' 2286 2103 Chain bag dredge. 750 kg total. Mostly pillow fragments with convex glass surfaces. A few bulbous pillow chunks, with cavities in cores. Some slabs, a few of which are glassy on both sides. Some breccia-encased fragments. Fresh olivine basalts, fine-grained and vesicular. Oxide coatings to 1 cm in thickness, but not ubiquitous. One angular chunk of altered, medium-grained basalt. 7 70-16-5 Heck Seamount Chain: east peak, eastern flank near summit 48°20' 128°15' 1463 1426 Pipe dredge. 50 kg total. Mixture of foraminifer-rich mud, gravel-sized basalt fragments, glass fragments, clinker, and one holocrystalline basalt specimen (7-1) with alteration rim and scattered plagioclase pheno-crysts. Also coarse-grained basalt (7-5) and nodule in cryptocrystalline clinker fragment (7-6). S PZ69-1 Heck Seamount, north flank 48°25.8' 129°22.5' 1445 1335 Pipe dredge. 200 kg total. Heterogeneous mixture of Mn-impregnated sponges, mud and irregular glassy basalt fragments. Basalt is vesicular, cryptocrystalline. Outer glass is palagonitized. Basalt forms suggest aa and pahoehoe flows. Oxide coatings uncommon, but maximum observed thickness is 2 mm. 9 PZ69-2 Heck Seamount, south flank 48°24' 129°23' 1150 1135 Pipe dredge. 50 kg total. Muddy mixture containing tiny glass and basalt chips, and chunks and slabs of laminated aquagene tuff. 10 PZ69-3 Heck Seamount, from west to east across crestal region 48°24.5' 129°24' 1150 1130 Pipe dredge. 30 kg total Mixture of sponge fragments [Mn-impregnated), basaltic debris, tuffaceous material, and vari-coloured sedimentary deposits (altered tuffs ?). Plus one 3 kg piece of massive, glassy basalt (Spec. 10-1). TABLE XIII--Continued DREDGE L 0 C A T I 0 N DEPTH (m) STATION Feature N. Lat. W. Long. From To BRIEF DESCRIPTION AND REMARKS 11 70-16-9 Heck Seamount Chain: west peak, east flank 48° 32' 129° 32' 1830 1740 Pipe dredge. 25 kg total Irregular fragments of fine-grained basalt, some glassy, most with oxide coatings, up to 6 mm in thickness. Basalt is generally vesicular and in many cases altered. One Fe-Mn nodule was recovered (Spec. 11-5). 12 71-1S-10 Heckle Seamount Chain: east end, southwest flank 48° 09' 129° 42' 2660 2325 Chain bag dredge. 45 kg total. Fragments of large pillows, broken along radial jointing. All have glass selvages (altered to palagonite on exterior) and oxide coatings up to 5 mm in thickness. Basalt is massive and fine-grained. 13 71-23-5 Heckle Seamount Chain: west peak, southeast flank 48* 25.5' 130° 06.4' 2470 2105 Chain bag dredge. 6 kg total. One piece of fine-grained, non-porphyritic basalt with very narrow glass selvage at one end. Encased in oxide coating up to 10 mm in thickness. Basalt has dark altered rim around a light grey core. One slab of Fe-Mn pavement was also recovered. 14 70-31-1 Continental slope: west wall of canyon on upper slope 48° 44' 126° 27.5' Approx. 730 Pipe dredge. Mixture of soupy brown mud, blue-grey clay, and chips of pale, brownish-grey (when dry) mudstone, generally riddled with worm burrows. 15 PZ69-16 Continental slope: west flank of h i l l on lower slope 48° 12.51 126°20.3' 2010 1005 Pipe dredge. Grey clay, small rounded igneous and metamorphic pebbles (glacial erratics), a few pieces of mudstone with worm burrows, and several pieces of pale, brownish-grey siltstone. 16 PZ69-15 Continental slope: west flank of h i l l on lower slope 48° 11' 126° 21' 1370 1190 Pipe dredge. Blue-green clay, small, rounded igneous and meta-morphic pebbles (glacial erratics), brownish-grey mudstone, and several pieces of pale brownish-grey, homogeneous, well-lithified siltstone. 17 PZ69-14 Continental slope: oast flank of h i l l on lower, slope 48° 11' 126° 15' Approx. 1460 Pipe dredge. Mixture of green mud, grey clay, pebbles (glacial erratics), and brownish-grey mudstone with worm burrows. TABLE XIV. Description of dredge stations at which no samples of underlying bedrock were obtained IOUBC L O C A T I 0 N DEPTH (m) Dredge S i t e F e a t u r e N. L a t . W. Long. From To R E M A R K S 71 -15-•5 Juan de Fuca Ridge: E a s t R i d g e , west f l a n k 48°31' 128°19' 2505 2330 Chain bag dredge. No r e c o v e r y , except a smear o f green mud on the dredge. No b i t e s i n d i c a t e d on te n s i o m e t e r (no o u t c r o p encountered). 71 -23- 2 Juan de Fuca Ridge: E a s t R i d g e , west f l a n k 48031.6' 128°17' 2470 2305 Chain bag dredge. No r e c o v e r y . No b i t e s i n d i c a t e d on t e n s i o m e t e r (no outcrop encountered). 70 -16- 1 Juan de Fuca Ridge: M i d d l e R i d g e , lower west f l a n k 48°44' 128°33' Approx. 2140 Pipe dredge. Recovery: s o f t , green mud, c o n t a i n i n g Recent, upper lower b a t h y a l f o r a m i n i f e r s which i n d i c a t e no v e r t i c a l displacement (B. E. B. Cameron, w r i t t e n comm.). 71 -15- 2 Juan de Fuca Ridge: M i d d l e R i d g e , upper west f l a n k 48°44.7' 128°32.5' Approx. 1830 C h a i n bag dredge. No r e c o v e r y . The dredge was caught on bottom, the wi r e snapped, and the dredge was l o s t . 71 -15- 6 Juan de Fuca Ridge: M i d d l e R i d g e , e a s t f l a n k 48*44.7' 128°32' 2195 1900 C h a i n bag dredge. No r e c o v e r y , except a smear o f s o f t , g r e e n , f o r a m i n i f e r - r i c h mud. 70 -16- 8 Juan de Fuca Ridge: M i d d l e R i d g e , west f l a n k 48°26.5' 128°36' 2415 2325 Pipe dredge. Recovery: s t i f f grey mud and s o f t green mud. The l a t t e r c o n t a i n s Recent lower b a t h y a l f o r a m i n i f e r s which i n d i c a t e no v e r t i c a l displacement (B. E. B. Cameron, w r i t t e n comm.). 70 -16- 3 Juan de Fuca Ridge: West R i d g e , near c r e s t 48°40' 128°48' 2250 2195 Pipe dredge. No r e c o v e r y . The dredge became t a n g l e d i n the wi r e (probably because i t was lowered too q u i c k l y ) and dre d g i n g was i n e f f e c t i v e . 70 -16- 6 Juan de Fuca Ridge: West R i d g e , e a s t f l a n k near c r e s t 48°40' 128°48' Approx. 2195 P i p e dredge. Recovery: s o f t greenish-brown mud and s t i f f e r g r e e n i s h - g r e y mud. The l a t t e r c o n t a i n s Recent lower b a t h y a l f o r a m i n i f e r s which i n d i c a t e no v e r t i c a l displacement (B. E. B. Cameron, w r i t t e n comm.). TABLE XIV--Continued IOUBC L 0 C A T I 0 N DEPTH Cm) Dredge S i t e F e a t u r e N. L a t . W. Long From To R E M A R K S 70-16-7 Juan de Fuca Ridge: West R i d g e , west f l a n k 4 8°28 ' 1 2 8°5 7 . 5 ' Approx. 2195 Pipe dredge. Recovery: s o f t greenish-brown mud ( c o n t a i n i n g ecent upper lower b a t h y a l f o r a m i n i f e r s ) and s t i f f e r g r e e n i s h - g r e y mud ( c o n t a i n i n g Recent upper lower b a t h y a l f o r a m i n i f e r s ) . No v e r t i c a l d isplacement i s i n d i c a t e d (B. E. B. Cameron, w r i t t e n comm.). 70-31-5 Juan de Fuca Ridge: West Ridge, south-west f l a n k 4 8°24 ' 1 2 8°59' Approx. 2835 Pi p e dredge. No r e c o v e r y . O p e r a t i o n hampered by stormy weather. 70-31-6a Sovanco Ridge, southern f l a n k 4 8°5 0 . 4 ' 1 2 9°23 ' Approx. 2600 Pi p e dredge. No r e c o v e r y . Same weather as above. 70-31-6b Sovanco Ridge, southern f l a n k 4 8°50' 1 2 9°20 ' Approx. 2650 Pi p e Dredge. No r e c o v e r y . Weak l i n k broken. Same weather as above. 71-15-4 Sovanco Ridge, e a s t e r n f a u l t s c a r p 4 8°5 4 . 6 ' 1 2 8°5 1 . 7 ' 2506 2100 Chain bag dredge. No r e c o v e r y . 70-16-4a Juan de Fuca Ridge: Endeavour Trough, e a s t e r n f a u l t s c a r p 4 8°1 1 ' 1 2 8°53 ' Approx. 2570 Pi p e dredge. No r e c o v e r y , d e s p i t e good b i t e s i n d i c a t e d by the t e n s i o m e t e r . 70-16-4b Juan de Fuca Ridge: Endeavour Trough, e a s t e r n f a u l t s c a r p 4 8°1 1 ' 1 2 8°55 ' Same Pipe dredge. No r e c o v e r y . Same as above. 70-31-4 Juan de Fuca Ridge: Endeavour Trough, e a s t e r n f a u l t s c a r p 4 8°1 1 . 4 ' 128°55» 2560 2451 Pi p e dredge. No r e c o v e r y . Same as above. 71-15-7 Juan de Fuca Ridge: Endeavour Trough, e a s t e r n f a u l t s c a r p 4 801 4, 1 2 8°56' 2652 2469 Chain bag dredge. No r e c o v e r y . Same as above. TABLE XIV--Continued IOUBC L 0 C A T I 0 N DEPTH On) Dredge Site Feature N. Lat. W. Long. From To REMARKS 71-23-4 Juan de Fuca Ridge: Endeavour Trough, eastern fault scarp 48°12" 128°S6' Approx. 2650 Chain bag dredge. No recovery. Same as above. 71-15-9 Heck Seamount Chain: western peak, northern flank 48°36' 129°46' 2705 2425 Chain bag dredge. No recovery. 71-23-6 Heck Seamount Chain: western peak, southern flank 48°33' 129°49' 2377 2195 Chain bag dredge. No recovery. Good bites were indicated on the tensiometer. The dredge returned battered around the mouth and smeared with green foraminifer-rich mud mixed with black Mn deposits'. 70-31-2a Upper continental slope 48°57' 126°31' 512 183 Pipe dredge. No recovery. Broken shear pin. 70-31-2b Upper continental slope 48°56' 126°32' Same Pipe dredge. Recovery: although the shear pin was broken, stiff grey-green and green clay was stuck to the inside of the dredge. Both types contain Pleistocene to Recent foraminifers which indicate no vertical displacement (B. E. B. Cameron, written comm.). 70-31-3 Upper continental slope 48°53' 126°38' Approx. 950 Pipe dredge. Recovery: green-grey mud whicli contains Pleistocene to Recent foraminifers. The sample was dredged from slightly deeper than the fauna indicates (B. E. B. Cameron, written comm.). 71-15-11 Lower Continental slope 48°35.6' 127°00' Approx. 2200 Chain bag dredge. No recovery. 268 APPENDIX B PART 2 WHOLE-ROCK CHEMICAL ANALYSIS OF BASALT BY ATOMIC ABSORPTION AFTER FUSION WITH LITHIUM METABORATE A rapid (?) procedure was devised for the determination of the major constituents of basaltic (and other silicate) rocks. Samples were fused with lithium metaborate, dissolved in dilute nitric acid, and diluted to appropriate concentrations. Aliquots of sample solutions and standard solu-tions were aspirated in air-acetylene or nitrous oxide-acetylene flames and atomic absorptions measured with a Techtron AA-4 spectrophotometer. S i , Al, T i , Fe, Mg, Ca, Na, K, and Mn were determined. Sr was attempted but the recommended value for the USGS standard basalt BCR-1 (Flanagan, 1967) was not obtained. Most of the method (described in detail here) is the result of synthesis of various aspects of techniques suggested by Suhr and Ingamells (1966), Langmyhr and Paus (1968), Medlin et al (1969), and Fletcher (1970). Interferences in the determinations of Ca and Mg caused by complexing with s i l i c a , alumina, sulphate, and phosphate are supressed by addition of lanthanum to sample and standard solutions. Ionization interferences between K and Na are supressed by addition of cesium to sample and standard solutions. Difficulties encountered in the determinations of Si and Al may have been partly due to interferences which are poorly documented for those elements. 269 The author wishes to thank Dr. E. G r i l l , Institute of Oceanography, U. B. C., for his assistance and advice, and for permitting the use of his laboratory and atomic absorption facilities for this work. Preparation of Reagents and Standards Reagents i . Nitric acid: 70-71% A. R., diluted to about 3.5%. i i . Cesium solution (10 mg Cs20/ml): transfer 11.95 g cesium chloride (Spex pure) to a 1 L volumetric flask, dissolve, and dilute to volume with water. i i i . Lanthanum solution (50 mg La/ml): suspend 58.7 g of lanthanum oxide (99.99%) in 250 ml water, cover, and add 250 ml hydrochloric acid. Heat to dissolve. Cool and dilute to 1 L with water, iv. Lithium metaborate, anhydrous (LiB02+): use directly in fusion. Also prepare concentrated solution in nitric acid (diluted) for use in preparation of secondary standard solutions. Primary Standards The methods of preparation of standard solutions for iron (0.1 mg/ml Fe203), magnesium (0.1 mg/ml MgO), calcium (0.1 mg/ml CaO), sodium (0.1 mg/ml Na20), and potassium (0.1 mg/ml K2O) are given by Fletcher (1970). Standard solutions for titanium (0.1 mg/ml Ti02) and manganese (0.1 ug/ml Mn) were obtained from Dr. E. G r i l l , Institute of Oceanography, U. B. C. A silicon standard solution was initially prepared by the procedure described by Langmyhr and Paus (1968). However, the recommended value for Si02 in the USGS standard basalt BCR-1 was not obtained using this standard. This was probably because the standard was not prepared with sufficient 270 accuracy in that the required ignition of silicon dioxide to constant weight at 1150 ± 50°C was uncertain. Consequently, a prepared standard containing 1 mg/ml Si was purchased from Harleco Laboratories. Aluminum (2 mg/ml AI2O3): dissolve 4.649 g of aluminum potassium sulfate in exactly 250 ml of water and HC1 (about 2 ml of HC1 in 248 ml of water). Secondary Standards A series of secondary.standard solutions is prepared for each oxide to cover the range of values likely to be encountered in the samples. This requires some educated guesswork, and i t may prove necessary to prepare additional standards after analysis begins. Secondary standard solutions should contain the same additives (La, Cs, and lithium metaborate) in approximately the same concentrations as the sample solutions. Because the La and Cs apparently adequately suppress interferences between elements, it does not appear to be necessary for the standard solutions to contain the same elements as the sample solutions. However, the presence of the additives themselves affects the absorbances of the elements, and hence, they must be present in both sample and standard solutions. 1. Oxides determined on the 1000 ug/ml sample solution (containing 8000 ug/ml lithium metaborate and 0.1 ml/ml La solution. If SrO is not being attempted, then the La solution may not be necessary) i . T i 0 2 : Range in samples 0.5-3% Sample solutions contain 5-30 ug/ml T i02- Hence, standard solution is diluted to 5, 10, 20, and 40 ug/ml T i 0 2 , with 8000 ug/ml lithium metaborate and 0.1 ml/ml La solution added. 271 i i . MnO: Range in samples 0.1-0.2% Sample solutions contain 1-2 ug/ml MnO. Hence, standard solution is diluted to 0.5, 1, and 2 ug/ml Mn (x 1.29 for MnO), with 8000 ug/ml lithium metaborate and 0.1 ml/ml La solution added. 2. Oxides determined on 400 ug/ml sample solutions (containing 3200 ug/ml lithium metaborate and 0.1 ml/ml Cs solution) i . Si02: Approximate value in samples = 50% Sample solutions contain about 100 ug/ml Si. Hence, the primary Si standard is diluted to 75, 100, and 125 ug/ml Si (x 2.14 for Si02), with 3200 ug/ml lithium metaborate and 0.1 ml/ml Cs solution added. i i . AI2O3: Approximate range in samples = 13-25% Sample solutions contain about 50-100 ug/ml AI2O3. Hence, the primary standard is diluted to 40, 60, 80, 100, and 120 ug/ml AI2O3, with 3200 ug/ml lithium metaborate and 0.1 ml/ml Cs solution added. i i i . K20: Approximate range in samples = 0-1% Sample solutions contain about 0-0.5 ug/ml K20. Hence, the primary standard is diluted to 0.25, 0.5, and 1 ug/ml K2O, with 3200 ug/ml lithium metaborate and 0.1 ml/ml Cs solution. A blank containing only the lithium metaborate and Cs solution is required for samples with less than 0.0625% K2O. 3. Oxides determined on the 50 ug/ml sample solution (containing 400 ug/ml lithium metaborate and 0.1 ml/ml La solution) i . Fe2Q3 (total iron as Fe203): Range in samples 7-12% Sample solutions contain 3.5-6.0 ug/ml Fe203- Hence, primary standard is diluted to 2, 4, 6, 8 ug/ml Fe203, with 400 ug/ml lithium metaborate and 0.1 ml/ml La solution added. 272 i i . CaO: Range in samples 10-13% Sample solutions contain 5-6.5 ug/ml CaO. Hence, primary standard is diluted to 4, 6, and 8 ug/ml CaO, with 400 ug/ml lithium metaborate and 0.1 ml/ml La solution added. 4. Oxides determined on the 5 ug/ml sample solution (containing 40 ug/ml lithium metaborate, 0.1 ml/ml La solution, and 10 ml/100 ml Cs solution) i . MgO: Range in samples 6-12% Sample solutions contain 0.3-0.6 ug/ml MgO. Hence, primary standard is diluted to 0.2, 0.4, and 0.6 ug/ml MgO, with 40 ug/ml lithium metaborate, 0.1 ml/ml La solution, and 0.1 ml/ml Cs solution added. i i . Na20: Range in samples 1.8-5% Sample solutions contain 0.09-0.25 ug/ml Na20. Hence, primary standard is diluted to 0.05, 0.1, 0.2, and 0.3 ug/ml Na20, with 40 ug/ml lithium metaborate, 0.1 ml/ml La solution, and 0.1 ml/ml Cs solution added. Preparation of Samples Break basalt specimens with a steel hammer on a steel plate and put fragments through a jaw crusher. From the resulting fragments, select chips to represent interiors, altered rims, or glass selvages as desired. Wash these chips in de-ionized water and dry at about 70°C. Grind chips in tungsten-carbide Spex ball mill to -100 mesh or less. Preparation of Sample Solutions Weigh 0.1 g of sample (ground to -100 mesh, or less) and 0.8 g lithium metaborate into a plastic weighing boat, mix well, and carefully transfer to a pre-ignited graphite cruciable. Place in muffle furnace at 1000°C for 15 minutes. 273 Add 50 ml of 3.5% nitric acid to a plastic beaker containing a magnetic stir bar. Remove cruciable from furnace, swirl to gather uncoalesced beads of molten material, and pour melt into the beaker. Cover to prevent evaporation loss and put on magnetic stirrer for at least one half hour. Pour into 100 ml volumetric flask, add 10 ml La solution, and dilute to 100 ml with washings (graphitic material will probably remain undissolved in the sample solution, but apparently does not interfere with analyses). If SrO is not to be attempted, i t is probably unnecessary to add La solution at this stage. This is sample solution 1. Transfer to clean plastic bottle. From sample solution 1 (containing 1000 ug/ml sample, 8000 ug/ml lithium metaborate, and 0.1 ml/ml La solution) prepare the following series of sample solutions: Sample solution 2--400 ug/ml sample, 3200 ug/ml lithium metaborate, and 0.1 ml/ml cesium solution. Sample solution 3--50 ug/ml sample, 400 ug/ml lithium metaborate, and 0.1 ml/ml lanthanum solution. Sample solution 4--5 ug/ml sample, 40 ug/ml lithium metaborate, 0.1 ml/ml lanthanum solution and 0.1 ml/ml cesium solution. Analytical Procedure Equipment A Techtron AA-4 atomic absorption spectrophotometer with a Texas Instrument Recorder was used for a l l determinations. Operating conditions for each element are given in Table XV. Acetylene flow and burner height (and gain) were not constant from day to day. It is necessary for the operator to adjust these parameters, as well as burner position, to obtain optimum conditions for each run. 274 TABLE XV. Analytical conditions for atomic absorption measurements Element Wave-length (A) Gain Slit width u Burner height setting Hollow cathode lamp current (ma) Flame* Acet. flow Si 2516 6 100 7-9 15 N20-acet 8-9 Al 3093 6 100 7-10 11 N20-acet 10 Ti 3643 1 100 9 20 N20-acet 10 Fe 2482 8 25 4-5 10 air-acet 3 Mg 2852 4 25 5 3 air-acet 2.5 Ca 4226 5 25 5 10 air-acet 2.5 Na 5890 5 50 6 5 air-acet 1.5 K 7665 11 300 5 5 air-acet 1.5 Mn 2794 3 25 5 10 air-acet 2.5 *Air or nitrous oxide pressure = 15 psi 275 Procedure (The operating instructions for the Techtron AA-4 spectrophotometer as summarized by Fletcher (1970) were generally followed by this analyst.) Zero with de-ionized water. 1. Aspirate appropriate series of secondary standards 2. Aspirate sample solutions and contiguous standards (aspirate de-ionized water between each) in the following order: i . Zero with de-ionized water, lower standard, sample solution, upper standard i i . Zero with de-ionized water, upper standard, sample solution, lower standard i i i . As for i iv. As for i i For each series, the content C in weight percent of the sample is calculated by the following equation: C = A + K where A is the weight % in the lower standard, K is the difference in weight % between the upper and lower standards, and E , E^, and are the absorbance of the sample solution, lower standard, and upper standard respectively. The arithmetic mean of the four series of measurements is finally calculated. The following flow chart summarizes the sample solutions which are required for the analysis of an ocean ridge basalt of the type found in the study area. Modification may be required for other basalts, and for other rock types. 276 Fuse (Ti02, MnO, SrO?) 20 ml of 1 + 5 ml Cs solution, to vol. in 50 ml vol. flask 0.1 g sample 0.8 g lithium metaborate + 50 ml nitric acid, 10 ml La sol., 40 ml washings, to vol. in 100 ml vol. flask Sample Solution 1 1000 ug/ml sample 8000 ug/ml Li Met. 0.1 ml/ml La sol. 5 ml of 1 + 10 ml La solution, to vol. in 100 ml vol. flask 02, K20, A1203) Sample Solution 2 400 ug/ml sample 3200 ug/ml Li Met. 0.1 ml/ml Cs sol. Sample Solution 3 50 ug/ml sample 400 ug/ml Li Met. 0.1 ml/ml La sol. 5 ml of 3 + 5 ml of La solution and 5 ml Cs solution, to vol. in 50 ml vol. flask Sample Solution 4 5 ug/ml sample 40 ug/ml Li Met. 0.1 ml/ml La sol. 0.1 ml/ml Cs sol. (Fe203, CaO) (MgO, Na20) 277 Accuracy and Precision of Chemical Analyses The samples were analyzed in four separate batches (referred to as 1, 2, 3, and 4) over a four-month period. In batches 1, 3, and 4, the USGS standard basalt BCR-1 was included. In batch 2, two samples from the Dellwood Seamount area which had been analyzed by classical wet chemical techniques by a commercial laboratory in Japan (Bertrand, 1972) were analyzed. In addition, subsequent to the completion of all the analyses, two samples from batch 3 were analyzed by the Geological Survey of Canada, Rapid Methods Analytical Group, Ottawa. In Table XVI are presented the comparative results of these analyses, which provide a measure of the accuracy of the lithium metaborate fusion-atomic absorption analysis technique developed by the author. In general, agreement is reasonable, and the variations observed are clearly not large enough to affect the conclusions drawn from the chemical analyses in this thesis. The variations do not suggest systematic error in the technique, and are probably to be expected in a routine and "rapid" analytical method. Some variations may be due to slight differences in the secondary standard solutions from one batch to another, or actual errors in the concentrations of the primary standard solutions. Also important are the operating conditions of the atomic absorption unit, particularly low amounts of acetylene and nitrous oxide in supply tanks which were observed to cause erratic A^O^ and Si02 readings. Similar problems have been encountered by other analysts (Medlin et_ al_, 1969). Atomic absorption analyses of major elements are apparently very sensitive to operating conditions in general (Dr. J . C. Van Loon, pers. comm., 1971). As mentioned in the description of the method, each sample with its contiguous standards is aspirated four times in succession, and the arithmetic 278 TABLE XVI. Comparisons between partial chemical analyses (Wt. %) of basalts analyzed by various labora-tories and by the author. Recommended values for BCR-1 are from Flanagan (1967). Mean deviations, of four aspirations are presented for each oxide of the author's analyses Recommended u.s G S. Standard Basalt BCR-1 Values Batch 1 Batch 3 Batch 4 Si02 54.48 54.3 + 0.3 54.5 ± 0.1 54.6 ± 0.2 A1203 13.65 13.3 0.1 13.5 0.2 13.0 0.2 Ti02 2.23 n.d, 2.24 0.05 2.13 0.02 Fe203 13.50 13.7 0.1 13.1 0.1 13.3 0.0 MgO 3.28 3.4 0.0 3.5 0.0 3.5 0.0 CaO 6.95 6.9 0.0 6.8 0.1 6.9 0.0 Na20 3.31 3.2 0.1 3.7 0.1 3.1 0.1 K20 1.68 n.d. n.d. n.d. MnO 0.17 0.19 0.0 0.18 0.0 0.18 0.0 70-025-2D 70-025-3D Japan Batch 2 Japan Batch 2 Si02 47.34 47.5 ± 0.2 50.04 49.9 ± 0.4 A1203 17.65 17.6 0.1 16.29 16.4 0.2 Ti02 1.24 1.13 0.04 2.37 2.32 0.03 Fe203 9.85 9.9 0.0 10.58 10.2 0.0 MgO 10.45 10.0 0.0 5.28 5.2 0.0 CaO 11.39 11.0 0.1 11.77 11.3 0.1 Na20 2.64 2.3 0.0 3.49 3.1 0.1 K20 0.16 0.15 0.0 0.52 0.41 0.00 Mno 0.14 0.16 0.0 0.15 0.16 0.00 8-3 1-4 GSC Batch 3 GSC Batch 3 s i o 2 49.5 49.0 ± 0.7 48.0 48.2 ± 0.7 A 1 2 ° 3 15.0 14.9 0.1 18.2 19.2 0.2 Ti02 1.00 0.84 0.03 0.89 0.82 0.02 Fe203 9.96 9.8 0.0 8.33 8.0 0.0 MgO 8.2 6.8 0.0 6.4 6.6 0.0 CaO 12.8 12.2 0.1 13.3 12.4 0.0 Na20 2.0 2.5 0.1 1.8 2.0 0.0 K20 0.2 0.06 0.00 0.1 0.09 0.0 MnO 0.16 0.16 0.00 0.13 0.13 0.0 279 mean of the four results is calculated. The mean deviations of each oxide are given in Table XVI. The maximum mean deviations obtained for the oxides of each batch are presented in Table XVII. These deviations generally reflect the stability of the cathode lamp for an element and the fact that the acetylene-nitrous oxide flame is less stable than the acetylene-air flame. To check the precision of the analyses, duplicate samples were run in batches 2, 3, and 4, and three samples were run in both batches 1 and 2. In addition, three samples from similar hand specimens (12-1, 12-2, 12-3) were analyzed, two in batch 4 and one in batch 3. These comparative analyses are presented in Table XVIII. Substantial agreement is observed, even between batches, indicating that precision is satisfactory. Further indications of the precision (and probably the accuracy) are provided by the notable similarity observed among analyses of fresh seamount basalts (Table IV), and the overall similarity of the analyses of basalts from the study area to analyses of ocean ridge basalts as a group (Fig. 40). In conclusion, although the analyses undoubtedly contain errors, these may be regarded as random and related to variables involved in the method, rather than to systematic errors. For the purposes for which they are required, the accuracy and precision of the chemical analyses are more than adequate. 280 TABLE XVII. Maximum mean deviations among the four aspirations of each sample recorded within each sample batch. The smallest standard deviations are associated with those elements whose cathode lamps are most stable, and whose solutions are aspirated in an air-acetylene flame Batch 1 Batch 2 Batch 3 Batch 4 Si02 ±0.44% ±0.43% ±0.92% ±0.61% A1203 0.23 0.20 0.21 0.30 Ti02 0.06 0.03 0.07 0.06 Fe203 0.09 0.08 0.12 0.17 MgO 0.04 0.07 0.05 0.07 CaO 0.06 0.06 0.11 0.11 Na20 0.07 0.05 0.09 0.07 K20 0.002 0.005 0.007 0.02 MnO 0.004 0.003 0.002 0.004 281 TABLE XVIII. Comparisons between replicate partial chemical analyses (Wt. %) in the four batches of samples. Mean deviations of four aspira-tions are also presented for each oxide of the author's analyses 9-1 11-1 Batch 1 Batch 2 Batch 1 Batch 2 Batch 2 - Dupl. Si02 41.9 + 0.2 42.8 ± 0.2 49.9 + 0.4 49.9 ± 0.4 50.8 ± 0.2 A1205 27.8 0.2 27.4 0.1 14.5 0.2 14.4 0.1 14.8 0.2 Ti02 0.80 0.01 0.83 0.0 0.96 0.02 0.93 0.03 0.93 0.03 Fe203 9.S 0.1 9.3 0.0 9.6 0.0 9.3 0.0 9.3 0.1 MgO 6.6 0.0 6.7 0.1 7.6 0.0 7.7 0.0 7.7 0.0 CaO 10.1 0.1 10.1 0.0 12.5 0.1 12.5 0.1 12.5 0.1 Na20 1.7 0.0 1.7 0.0 2.4 0.1 n.d. 2.2 0.1 K20 0.01 0.0 0.02 0.0 0.04 0.00 0.04 0.00 0.04 0.01 MnO 0.17 0.0 0.15 0.0 0.17 0.00 0.16 0.00 0.16 0.00 1-1 4-1 Batch 1 Batch 2 Batch 3 Batch 3 -Dupl. Si02 47.5 ± 0.1 47.3 ± 0.1 49.1 ± 0.5 49.1 ± 0.3 A1203 19.1 . 0.1 18.9 0.2 15.7 0.0 15.8 0.1 Ti02 0.93 0.02 0.73 0.01 0.72 0.04 1.03 0.01 Fe203 8.2 0.1 8.1 0.1 9.1 0.0 9.4 0.1 MgO 6.3 0.0 6.4 0.1 7.7 0.0 7.5 0.0 CaO 12.7 0.1 12.8 0.1 11.9 0.0 12.1 0.0 Na20 2.0 0.0 1.7 0.0 3.7 0.0 3.5 0.0 K20 0.08 0.00 0.08 0.00 0.03 0.00 0.05 0.01 Mno 0.13 0.00 0.12 0.00 0.16 0.00 0.16 0.00 5-1 12-2 12-3 12-1 Batch 4 Batch 4 - Dupl. Batch 3 Batch 3 Batch 4 Si02 48.4 + 0.2 48.7 ± 0.4 49.0 + 0.4 49.5 + 0.3 50.5 ± 0.1 A1203 19.9 0.2 19.3 0.2 15.7 0.2 14.7 0.1 14.6 0.1 Ti02 0.89 0.03 0.85 0.02 1.33 0.07 1.22 0.02 1.22 0.03 Fe203 8.5 0.1 7.8 0.0 8.9 0.0 9.1 0.1 9.2 0.1 MgO 5.9 0.0 7.1 0.1 7.0 0.0 7.2 0.1 7.0 0.1 CaO 11.8 0.1 12.5 0.1 11.6 0.1 11.5 0.0 11.6 0.0 Na20 2.1 0.0 1.9 0.0 2.6 0.1 2.5 0.0 2.1 0.0 K20 0.07 0.0 0.08 0.01 0.11 0.0 0.09 0.0 0.08 0.1 MnO 0.12 0.0 0.12 0.00 0.18 0.0 0.18 0.0 0.18 0.0 282 TABLE XIX. Petrographic descriptions of analyzed specimens Sample Number IOUBC Sample Number Petrographic Description 1-•1 70 -16- •2--7 15% plagioclase xenocrysts in fine-grained groundmass. Specimen is altered but freshest (grey) parts used for analysis. T.S.: Hyalopilitic to intersertal glomero-porphyritic basalt. Resorbed plagioclase xenocrysts An85. 2% interstitial green alteration product, also is pseudomorphic after scattered olivine phenocrysts. 1- 1AR 70 -16-AR 2-•7 Altered rim of specimen 1-1. Plagioclase crystals yellowish. Groundmass yellowish-brown. T.S.: Olive-green chlorophaeite (?) in vesicles and interstices. 1- 2 70 -16- 2- 5 Glomeroporphyritic plagioclase basalt like 1-1, but only 3% xenocrysts. T.S.: Like 1-1, but few plagioclase xenocrysts. 1- 3 70 -16- 2- 12 Glomeroporphyritic plagioclase basalt with 15% xenocrysts. Like specimen 1-1. No T.S. 1- 4 70 -16- 2- 19 Glomeroporphyritic plagioclase basalt with 15% xenocrysts. Like specimen 1-1. No T.S. 1- 5 70 -16- 2- 1 Glomeroporphyritic plagioclase basalt with 10% xenocrysts. T.S.: Like specimen 1-1 but fewer xenocrysts (Fig. 30b). 2- 1 71 -23- 1- 1 Angular piece of massive, f.g. grey basalt with darker altered rim. T.S.: Intergranular texture, with plagioclase laths (54%), clinopyroxene (40%), opaques (3%), and interstitial brown alteration products (3%) (Fig. 30e). 2- 2 71 -23- 1- 7 Fine-grained, minutely vesicular basalt encased in palagonite breccia. T.S.: Intersertal texture, in patches approaching intergranular where clinopyroxene and plagioclase are coarser. Unusually high concen-tration of opaque minerals (10-15%) (Fig. 30f). 2- 2AR 71 -23-AR 1- 7 Darker grey rim (1 cm. thick) around specimen 2-2. T.S.: Vesicles and interstices of intersertal basalt fi l l e d with yellow or reddish-yellow chlorophaeite (?) and iron oxides. 2- 3 71 -23- 1- 21 Vesicular, glassy basalt. T.S.: Glassy to hyalopilitic, with scattered plagioclase microlites (An 50-60). 283 TABLE XIX--Continued Sample Number IOUBC Sample Number Petrographic Description 3-1 71 -23 -3- 3 10% fresh olivine xenocrysts in glassy basalt. T.S.: Hyalopilitic texture. Euhedral to anhedral olivine xenocrysts (Fo 90). 15% plagioclase laths (An 70) (Fig. 31a, b). 3-1G1 71 -23 Gl -3- 3 Fresh, lustrous, black glass selvage of specimen 3-1. Contains olivine xenocrysts like the hyalopilitic interior (Fig. 31a). 4-1 71 -15 -8- 1 10% fresh plagioclase xenocrysts and rare olivine phenocrysts in glassy basalt. T.S.: Variolitic texture. Plagioclase xenocryst compositions An 85-90 (Fig. 31c, d, e). 4-1G1 71 -15 Gl -8- 1 Fresh, lustrous black glass selvage of specimen 4-1. Contains xenocrysts like the variolitic interior. 5-1 71 -15 -5- 15 15% plagioclase xenocrysts and rare olivine phenocrysts (altered) in fine-grained basalt with glass margin. Glomeroporphyritic. No T.S. Altered. Similar to basalts from Station 1. 5-2 71 -15 -5- 26 10% plagioclase xenocrysts in a basalt like specimen 5-1. No T.S. Altered. 5-3 71 -15 -5- 12 Angular piece of massive, holocrystalline, medium-grained basalt. T.S.: Intergranular texture, with plagioclase laths (40%), clinopyroxene (45%), and opaques (5%). 5-4 71 -15 -5- 13 Like specimen 5-3. T.S.: Like specimen 5-3. 5-5 71 -15 -5- 24 Fine-grained, minutely vesicular basalt with glass margin. No T.S. 6-1 71 -23 -7- 3 2-5% tiny vesicles in a fine-grained basalt (pillow fragment) with glass margin. T.S.: Texture hyalopi-l i t i c to intersertal, with 35% plagioclase microlites (An 65), 5% fresh olivine microphenocrysts, and 60% turbid, cryptocrystalline groundmass (Fig. 34f). 6-2 71 -23 -7- 6 Fine-grained, massive basalt with dark alteration rim. T.S.: Intersertal texture, with very fine-grained plagioclase needles and clinopyroxene, opaque minerals, and 5% olivine microphenocrysts (Fig. 34e). 284 TABLE XIX--Continued Sample Number IOUBC Sample Number Petrographic Description 7-1 70-16-5-8 Angular piece of massive, fine-grained basalt with dark altered rim. T.S.: Intergranular texture (holocry-stalline) with plagioclase, clinopyroxene, and opaques. Scattered plagioclase phenocrysts (An 70-80) (Fig., 35c, d). 7-1AR 70-16-5-8 AR Dark altered rim of specimen 7-1. T.S.: Green chlorophaeite (?) in interstices and vesicles of inter-granular basalt. 7-2 70-16-5-11 Cryptocrystalline clinker basalt, with round or tubular vesicles and scattered plagioclase phenocrysts. Altered. No T.S. 7-3 70-16-5-10 Cryptocrystalline clinker basalt, with scattered vesicles, and more plagioclase phenocrysts than specimen 7-2. Extensively altered. No T.S. 8-1 PZ69-1-1 Cryptocrystalline basalt with glass selvage, large, tubular vesicles, and scattered plagioclase pheno-crysts. No T.S. 8-1G1 PZ69-1-1 Gl Fresh, lustrous black glass selvage of specimen 1-1. 8-2 PZ69-1-2 Cryptocrystalline dark grey basalt, with scattered plagioclase and rare olivine phenocrysts. The specimen has a glass selvage at one end. T.S.: Variolitic texture, with plagioclase microlites in places forming the nuclei of the varioles. 8-3 PZ69-1-13 A large, relatively massive piece of cryptocrystalline basalt, with a narrow glass margin at one end. It contains scattered plagioclase and rare olivine pheno-crysts, which tend to cluster. Vesicles have concen-trated between glass and interior, and in core of specimen. No T.S. 9- 1G1 PZ69-2-3 A 2.5 cm X 2.5 cm slab of basaltic glass, with scattered small plagioclase phenocrysts. Outer glass is altered to yellow-brown palagonite, but this was avoided in sampling for chemical analysis. 10 -1 PZ69-3-1 Fine-grained grey basalt with glass selvage at one end, and narrow and patchy darker altered rim. T.S.: Intergranular texture with 45% palgioclase (An 60-70), 52% clinopyroxene, and 3% opaques. Scattered plagio-clase and olivine phenocrysts (Fig. 36a). 285 TABLE XIX--Continued Sample Number IOUBC Sample Number Petrographic Description 10- 1G1 PZ69-3 Gl -1 Fresh, lustrous black selvage of specimen 10-1. 11- 1 70 -16- 9-4 Fine-grained, massive basalt with rare plagioclase phenocrysts and a narrow dark grey altered rim. T.S.: Intergranular texture (holocrystalline) with plagioclase (An 55), clinopyroxene, and opaques. Scattered plagioclase phenocrysts (An 70), Similar to specimen 7-1. 11- 2 70 -16- 9-5 Cryptocrystalline basalt with (max. 3 mm) oxide coating. Vesicular. Extensively altered throughout in a swirly pattern of yellowish and brownish greys. Alteration in part follows fractures. No T.S. 11- 3 70 -16- 9-6 Vesicular, cryptocrystalline basalt with glass selvage and dark grey-brown altered rim. Specimen is altered throughout, but less than specimen 11-2. No T.S. 12- 1 71 -15- 10- 3 Pillow fragment of fine-grained basalt with thick (1 cm) glass selvage along curved surface and thick (up to 5 mm) oxide coating. Contains scattered plagio-clase phenocrysts. Fresh basalt is medium-grey, but dark grey alteration rims are extensively developed. T.S.: Intersertal to intergranular (in patches) with plagioclase laths, clinopyroxene (usually extremely fine-grained), and opaques (about 5%). 12- 1AR 71 -15-AR 10- 3 Dark altered rim of specimen 12-1. T.S.: Vesicles and interstices f i l l e d with yellow-green chlorophaeite (?) and iron oxides, which constitute about 15% of the basalt. 12- 2 71 -15- 10- 7 Similar to specimen 12-1. No T.S. 12- 2AR 71 -15-AR 10- 7 Similar to specimen 12-1AR. No T.S. 12- 3 71 -15- 10- 9 Similar to specimen 12-1. No T.S. 12- 3AR 71 -15-AR 10- 9 Similar to specimen 12-1AR. No T.S. 286 TABLE XIX--Continued Sample Number IOUBC Sample Number Petrographic Description 13-1 71-23-5-1 Fine-grained basalt, with minor glass at one end. The fragment is completely encased in oxide coating up to 1 cm in thickness. There is a wide dark alteration rim. Scattered small vesicles are empty in the fresh parts, but f i l l e d with green chlorophaeite (?) in the dark altered parts. T.S.: Intersertal to intergranu-lar (in patches) texture, like specimen 12-1 (Fig. 36e). FIGURE 15. Revised bathymetric chart of the northern end of Juan de j Fuca Ridge and adjacent continental s lope. Contour interval 100m. Depths in metres, corrected for sound velocity. r 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0052648/manifest

Comment

Related Items