Diabase Sheets Research Articles

Reconstruction of the geometry of accretion during formation of the Bay of Islands Ophiolite Complex

Two types of highly allochthonous ophiolitic terrains crop out adjacent to one another in the Bay of Islands region of southwestern Newfoundland, Canada. The first, called the Coastal Complex (CC), is exposed as a long, linear belt along the west coast. It consists of a highly deformed ophiolitic assemblage interpreted as oceanic crust and upper mantle with a previous transform fault tectonic history. The second, called the Bay of Islands Complex (BOIC), crops out to the east and consists of a series of four relatively undeformed ophiolitic massifs interpreted as oceanic crust and upper mantle with a more normal seafloor spreading tectonic history. An autochthonous relationship can be demonstrated between the two types of ophiolitic terrains in the southernmost of these massifs, indicating that each evolved along a single ridge‐transform system. The internal structure of both the BOIC and the CC is consistent with the above interpretations. The NE‐SW structural trends in the CC transform are oriented approximately parallel to the sea‐floor spreading direction inferred from the average orientation of the sheeted diabase dikes within the BOIC. One‐way chilling statistics from the sheeted diabase dike unit, growth directions of dendritic crystals along inclined comb layers in the plutonic section, and the sense of shear along the CC transform all indicate that the BOIC was accreted on the NE side of a NW‐SE trending spreading center (present geographic reference frame). Juxtaposition of the BOIC with the NE‐SW trending CC occurred at a ridge‐transform intersection. The transform exhibited a left‐lateral offset. Magma chamber profiles have been identified in the layered gabbroic sections of the BOIC by using the patterns outlined by the attitudes of fine‐scale igneous layering. These patterns depict arcuate structures which appear to define roughly a cross‐sectional view of a systematically oriented half‐cylinder; the long axis of which is oriented parallel with the plane of the sheeted dikes and major lithologic unit contacts. Thus we believe that a large cylindrical‐shaped crustal level subaxial magma chamber extended in a direction parallel with the ridge axial‐plane at the time of formation of the BOlC. The CC marks the abrupt truncation of this long, linear magma chamber. Petrologic constraints demonstrate that the ultramafic and transition zone cumulates which form the base of the plutonic section, crystallized at a range of high to intermediate pressures possibly along a long narrow conduit extending to 30‐km depth beneath the crustal level magma chamber. After formation, these cumulates were deformed at high temperatures in the zone of mantle upwelling and divergence, then tectonically transported to crustal or near crustal levels to form the basement upon which the remainder of the ophiolite stratigraphy was accreted. A geometric and tectonic model is presented for the accretion process responsible for formation of the Bay of Islands Ophiolite.

Age and stratigraphic–tectonic significance of Proterozoic diabase sheets, Mackenzie Mountains, northwestern Canada

Diabase dikes and sills were intruded during a Late Proterozoic phase of extensional tectonics in the northern Canadian Cordillera. This event is strongly expressed in the depositional pattern of formations immediately below and above the base of the Windermere Supergroup. Rb–Sr isochron dates of 766 ± 24 and 769 ± 27 Ma have been obtained for two of the these diabase sheets, which intruded the Tsezotene Formation of the Mackenzie Mountains Supergroup. The emplacement of the diabase sills and dikes may be related to either of two volcanic events in the region, one documented, the other inferred. In either case, emplacement of the diabase at approximately 770 Ma is close to the boundary between the Mackenzie Mountains Supergroup and the overlying Windermere Supergroup, precedes a regional glaciation (Shezal tillite), and reflects a strong accentuation of the west-facing Cordilleran miogeocline.

Geologic section through the Samail Ophiolite and associated rocks along a Muscat‐Ibra Transect, southeastern Oman Mountains

Regional mapping at a 1 : 60,000 scale of a 30‐km strip from the Gulf of Oman (Muscat) across the Oman Mountains, 130 km to the south, provides the geologic setting for the (∼95 m.y.) Ibra section of the Samail ophiolite. Where best preserved, the Ibra ophiolite section is an ∼8 km‐thick section of oceanic consisting of ∼0.5 km of pillow lavas, 1.2–1.6 km of sheeted diabase dike complex, 0.2–1.0 km of high‐level noncumulate gabbro, and 3.0–5.0 km of cumulate gabbro that is underlain by tectonite peridotite 9–12 km thick. The Ibra section is found on the southward dipping limb of the Sayah Hatat antiform. The tectonite peridotite represents uniformly depleted harzburgite and dunite that have been deformed by high‐temperature, low‐stress asthenospheric flow. Discordant dunites within the tectonite peridotite appear to represent either flow crystallization products from primary picritic liquids or reaction products of these liquids with the harzburgite. The structural base of the tectonite peridotite is overprinted by a high‐stress, low‐temperature deformation that can be related to its oceanic detachment. The layered gabbros are predominantly olivine‐clinopyroxene‐plagioclase cumulates, and orthopyroxene does not occur as a cumulus phase. Occurrence of cumulate wehrlites and picrites at high stratigraphic levels within the layered gabbros is evidence that the gabbroic section crystallized predominantly from the bottom upward in a periodically replenished magma chamber. High‐level gabbro represents remnants of crystallization at the roof of the magma chamber and intrudes most overlying diabase dikes. Both the diabase dike complex and pillow lavas are hydrothermally altered, and alteration and metamorphism increase downward (zeolite (?) to epidote‐amphibolite facies). In spite of pervasive alteration, relict primary mineralogy and bulk chemistry suggest that the diabase dikes and pillow lavas are cogenetic with the underlying gabbros. The present‐day Samail thrust surface truncates ophiolite stratigraphy and puts the Samail ophiolite on top of unmetamorphosed Hawasina melange. The last motion on this surface was probably no older than Maestrichtian (70–65 m.y.). Garnet amphibolites exposed as remnants of earlier thrusting (∼90 m.y.) record initial ophiolite detachment at a 14–20‐km depth within the Tethyan oceanic lithosphere. The Hawasina Group, which underlays the Samail ophiolite, is a block melange where exposed near Muscat in the north, and it grades into an imbricated broken formation at the southern limit of the map area. The Hawasina Group is thrust over Permian to Late Cretaceous shelf carbonates that represent autochthonous Arabian continental shelf deposits. Recent (post‐Miocene) collapse and dome structures, such as the Ibra dome, have complicated ophiolite stratigraphy south of Jabal Dimh. Our geologic studies strongly indicate the Samail ophiolite represents a large, coherent slab of transported oceanic lithosphere formed at a Late Cretaceous spreading center in the Tethyan Sea.

Deformation of the rock foundation of the Bratsk hydroelectric station

Deformations of the rock foundation of the Bratsk dam occurred mainly during filling of the reservoir and primarily due to the weight of the water operation. During operation, settlements have increased insignificantly with a clear tendency toward their fading away. The maximum settlement, 74 mm, which had been predicted as early as 1966, was reached in 1970; in 1977 it amounted to 74.4 mm. Horizontal deformations have practically stabilized. At the base of the dam, if slight consolidation occurred at all during construction (which, apparently, caused mainly differential settlement of the sections), then it was hardly noted by observations. Despite the explicit stabilization of the processes, observations of settlements and horizontal displacements retain their importance, considering the essential character of the structures, the seismicity of the region, and the slowness of processes characteristic for hydraulic structures. The geodetic base of the Bratsk hydrostation has certain shortcomings: the bench marks are located on a diabase sheet and can settle along with it; the nearness of the diabase rock quarry to the right-bank bench marks; the rather long and complex traverse along the foreshores. Use of the geodetic base is also difficult: the traverses along the foreshores are difficult and dangerous owing to the adjacent high (up to 80 m) steep cliffs, the natural disintegration of which causes rock falls. We consider that it would be more convenient and reliable to embed bench marks with anchors in sandstones below the upper zone of consolidation (depth 100–120 m) in the bank tunnels of the inspection gallery at the 27-m level.

A preferred orientation of intrusion of Precambrian dykes

AN analysis of the original orientation of Precambrian dykes has produced a nonrandom pattern that implies the existence of a Precambrian stress field of global extent. In general, dyke intrusion tends to occur in a north–south orientation, and is concentrated in low latitudes. This distribution has many similarities to the present distribution of sheeted diabase complexes of present day oceanic ridge crests, and may reflect the persistence of the same stress field throughout geological time. Other evidence in favour of the existence of this stress field can be found in tectonic motions or geological features that exhibit a symmetrical distribution relative to the Earth's rotational axis. For example, Moore1 showed that the preferred orientation of spreading axes is directed north–south with a secondary maximum orientated east–west. Active transform faults exhibit a unimodal east–west distribution. A former study of recent continental rift valleys has revealed a preferred north–south orientation of fracturing. Cainozoic grabens, most of which have not been significantly rotated by plate movements following their formation, exhibit an especially strong peak within 10° of north2. Here we describe our analysis of preferred fracture orientations using the trends and directions of Precambrian dykes. A central point of this study is the assumption that regional dyke patterns record fractures which result from regional stress fields operative at the time of dyke intrusion. Escher et al.3 have confirmed the above assumption using geological evidence from the Amitsoq dykes of Greenland.

The Geology and Geochemistry of Ophiolitic Rocks Exposed at Ming's Bight, Newfoundland

The Baie Verte Group, as exposed on the peninsula between Baie Verte and Ming's Bight, consists of an ophiolite assemblage ranging from interlayered ultramafic and gabbroic rocks to sheeted diabase dikes overlain by pillow lavas and volcanic sediments. The sequence has been disrupted into five structural blocks separated by fault zones containing serpentinized peridotite and/or talc-carbonate; units within each block are separated by less significant faults. These structures and other deformation features in the Baie Verte Group are interpreted to be related to early Ordovician emplacement with some effects of later Acadian deformation.The Baie Verte Group is chemically similar, both in major and trace elements, to other ophiolite sequences such as in Oman and Papua. A low-Ti and low-K tholeiitic magma crystallized under conditions of low oxygen fugacity in the upper crust beneath a mid-ocean ridge, producing the observed peridotite–pyroxenite–gabbro–diabase–pillow lava sequence.

Crystallization history of a zoned plagioclase

SummaryElectron microprobe and textural data are used to deduce a crystallization history for plagioclase in the Grand Manan Island (Canada) diabase sheet. The plagioclase crystals consist of a prominent core (An79-67) surrounded by a distinctly more calcic zone (An82-76), which grades into a normally zoned mantle (An73-40). This zoning pattern is interpreted as reflecting the elevation of a dry diabase magma from at least 10 to 20 km to near surface conditions.

Lead Isotope Evidence on the Genesis of the Silver-Arsenide Vein Deposits of the Cobalt and Great Bear Lake Areas, Canada

Silver veins of Proterozoic age in the Cobalt area, Ontario, are associated with the thick Nipissing diabase sheet, and those at Great Bear Lake, N.W.T., are also, in part, associated with a diabase sheet. Host rocks in the Cobalt area are flat-lying Aphebian sedimentary rocks, steeply dipping Archean volcanic and associated sedimentary rocks, and diabase, while host rocks in the Great Bear Lake area are intermediate and felsic volcanic rocks and tuffs of the Echo Bay group.Model-lead ages for ordinary galenas are about 1,630 and 2,280 m.y. for the Great Bear Lake and Cobalt veins, respectively, whereas the pitchblende U-Pb age for the Great Bear Lake veins is about 1,445 m.y., and the apparent Rb-Sr age for the Cobalt veins is about 2,160 m.y. An Rb-Sr isochron age of 1,425 + or - 48 m.y. has been obtained for the diabase sill in the Port Radium area, Great Bear Lake.In the Cobalt area essentially identical lead isotope compositions were obtained for interflow base metal mineralization in Archean volcanic rocks, galena along bedding in sedimentary rocks of the Cobalt group, and some late-stage sulfide veins. One interflow sample with a more primitive composition has a model age of at least 2,805 m.y. Mineralization, including deposition of most of the lead in Archean interflow beds, is interpreted to have taken place at about 2,160 m.y. and to be genetically related to the Nipissing diabase. In addition, two anomalous lead lines are defined, one of which suggests an age of 3,260 + or - 100 m.y. for Archean volcanic rocks. A shallow line, represented in only late-stage sulfide veins, is difficult to interpret but could be due to the late addition of a radiogenic component during Paleozoic or younger events not normally considered as causing mineralization, or even from ground water.Leads from the Great Bear Lake area could represent a single anomalous lead line with a slope of 0.1085 + or - 0.0070, and thus have a maximum possible age of 1,130 m.y. Geological events younger than this are known in the Bear Province, but not in the vicinity of the veins, and the same problems of interpretation exist as for the Cobalt area.The source of lead is not defined in either area. However, a source with homogeneous lead is required and the lack of extensive wall-rock alteration suggests that the lead was not leached from the country rocks. The diabase is tentatively favored as the source, but a connate brine or other source is possible.

1539. Measurements of the energy distribution of backscattered kilovolt electrons with a spherical retarding-field energy analyser: T Matsukawa et al,J Phys D: Appl Phys, 7 (5), 1970, 695–702

Recently discovered horizons of non-deformed felsic welded crystal tuff in the Proterozoic rift system of northern Ontario yield a geologically stable paleomagnetic vector. Using alternating field (AF), thermal (TD) and low-temperature demagnetization (LTD) techniques, a single stable characteristic remanence (ChRM) vector is found in all specimens (declination/inclination = 115°/−73.5°; α95 = 2.1°, n = 58). It is overprinted by a Bruhnes’ epoch VRM. The tuffs paleopole position coincides with that of ubiquitous sub-horizontal diabase sheets which intruded at ∼1110 Ma during Late Proterozoic Keewanawan Volcanism. The latter produced numerous volcanic and intrusive rocks along the margins of the Lake Superior paleo-rift system, during the interval 1087–1120 Ma, but the sills are most extensive. Although the Voosges Lake Tuff shares the same paleopole location as the diabase sheets (1109 Ma), the nearby Seagull pluton (1112 Ma) and the Keweenawan volcanic rocks, its primary zircons yielded a Mesoproterozoic crystallization age of 1542 ± 3 Ma. This is similar to the nearby English Bay felsic intrusions dated at 1537−2+10Ma [Davis, D.W., Sutcliffe, R.H., 1985. U-Pb ages from the Nipigon “Plate” (sic) and northern Lake Superior. Geol. Soc. Am. Bull. 96, 1572–1579]. Whereas the paleomagnetic magnetic signal appears primary in the 1542 Ma Voosges Tuff, it was completely reset by the Late Proterozoic Keewanawan paleofield (∼1100 Ma). The Tuff units are separated from the diabase sills by stratigraphic thicknesses of ≤50 m and show no traditional petrographic evidence of annealing or thermal metamorphism. Since it is improbable that the 1542 Ma tuff returned to the same paleolatitude at 1109 Ma, we postulate that metasomatic processes caused a ∼1109 Ma remagnetization. A similar event could explain unusual ChRMs reported for certain other pre-Keweenawan Proterozoic sites with adjacent Keewanawan diabase sheets. For example, some paleopoles reported for the Gunflint Formation (1878 Ma) and Sibley Formation (>1339 Ma, perhaps ∼1537 Ma by association with the English Bay intrusions) might have been acquired due to secondary remagnetization.

Lushs Bight and Roberts Arm Groups of Central Newfoundland: Possible Juxtaposed Oceanic and Island-Arc Volcanic Suites

The Lushs Bight Group in the Pilleys Island area of central Newfoundland consists of basaltic pyroclastic rocks, pillow lavas, and sheeted diabase dikes, cut by plugs of gabbro and quartz diorite which have produced local amphibolitization. The Lushs Bight rocks are commonly gently folded but locally strongly schistose along shear belts and fault zones and around the intrusive bodies. The sheeted diabase has been metamorphosed to the upper greenschist (actinolite) facies, while in the flows, chlorite is dominant over actinolite, indicating a lower grade of metamorphism. The Roberts Arm Group consists of basaltic pillow lava locally overlain by dacite representing volcanic centers. This group has a strong bimodal chemistry; it differs from the Lushs Bight Group in the presence of dacite, in a lack of actinolite-grade metamorphism, and in that the basalt of the Roberts Arm Group is richer in K 2 O, Na 2 O, TiO 2 , Ba, Rb, and Zr, and poorer in CaO. These data are taken as evidence that the Roberts Arm volcanic rocks possibly represent an island-arc suite which was deposited upon oceanic crust now represented by the Lushs Bight Group.

Favorability for Cornwall-type magnetite deposits in Pennsylvania using geological, geochemical and geophysical data in a discriminant function

Up to 18 geological, geochemical and geophysical variables are combined in discriminant functions to evaluate the favorability of the Triassic province in Pennsylvania for Cornwall-type magnetite deposits. Variables include regional stream sediment data, percentages of various rock types in the drainage, aeromagnetic data, and estimates of the thickness of diabase sheets and extent of diabase-carbonate contacts. Both a simple discriminant function and a stepwise discriminant method give similar patterns of favorable areas, as does analysis of factor scores after an R-mode factor analysis. Results are slightly improved by correcting chemical data for variable background. Despite its semi-quantitative nature, the extent of diabase-carbonate contacts is found to be the most important variable, followed by magnetic intensity, thickness of diabase, and one or more chemical parameters. In addition to the known areas of mines and prospects, additional favorable areas near Boyertown and Gettysburg are recognized from the discriminant treatment.

Permian-Triassic Beacon Group of the Shackleton Glacier Area, Queen Maud Range, Transantarctic Mountains, Antarctica

Beacon Group sediments, consisting of approximately 1425 m of Permian-Triassic graywacke sandstones, siltstones, shales, and coals, are well exposed in the Shackleton Glacier area. The strata have a terrestrial origin and are subdivided into five formations. In ascending order, the formations are Pagoda Formation, Mackellar Formation, Mt. Butters Formation (new name), Buckley Formation, and Mt. Kenyon Formation (new name). Numerous Jurassic diabase sheets, individually as thick as 300 m, have an aggregate thickness about equal to that of the sedimentary section. The rock succession is similar to Beacon and diabase rocks in adjacent areas along 800 km of the Transantarctic Mountains and Gondwana rocks of other southern continents.

Paleomagnetic evidence for the extent of Mackenzie igneous events

North-northwesterly striking Mackenzie diabase dikes of middle-Proterozoic (Helikian) age are profuse in the western part of the Canadian Shield. Published paleomagnetic data on dikes of this trend in Mackenzie District, on the Muskox Intrusion, the Coppermine River volcanic rocks, and the Sudbury dikes suggest that they are all products of closely related igneous events. This paper presents paleomagnetic data that suggest that the intrusion of extensive diabase sheets in the East Arm of Great Slave Lake, and of dikes as far to the northeast as Melville Peninsula and as far to the southeast as Manitoba, were also parts of these events. The mean paleomagnetic pole position for the Mackenzie dikes and for related intrusive and extrusive rocks is [Formula: see text], 171 °W with [Formula: see text]. Radioactive age determinations, some of which are unpublished, indicate an age of about 1200 m.y. for the formation of these rocks. It is suggested that for convenience all of these apparently related intrusive and extrusive igneous episodes be referred to as Mackenzie igneous events.

Tectonics of Antarctica

Antarctica consists of large and wholly continental east Antarctica and smaller west Antarctica which would form large and small islands, even after isostatic rebound, if its ice cap were melted. Most of east Antarctica is a Precambrian Shield, in much of which charnockites are characteristic. The high Transantarctic Mountains, along the Ross and Weddell Seas, largely follow a geosyncline of Upper Precambrian sedimentary rocks that were deformed, metamorphosed and intruded by granitic rocks during Late Cambrian or Early Ordovician time. The rocks of the orogen were peneplained, then covered by thin and mostly continental Devonian-Jurassic sediments, which were intruded by Jurassic diabase sheets and overlain by plateau-forming tholeiites. Late Cenozoic doming and block-faulting have raised the present high mountains. Northeastern Victoria Land, the end of the Transantarctic Mountains south of New Zealand, preserves part of a Middle Paleozoic orogen. Clastic strata laid unconformably upon the Lower Paleozoic plutonic complex were metamorphosed at low grade, highly deformed and intruded by Late Devonian or Early Carboniferous granodiorites. The overlying Triassic continental sedimentary rocks have been broadly folded and normal-faulted. Interior west Antarctica is composed of miogeosynclinal clastic and subordinate carbonate rocks which span the Paleozoic Era and which were deformed, metamorphosed at generally low grade, and intruded by granitic rocks during Early Mesozoic time and possibly during other times also. Patterns of orogenic belts, if systematic, cannot yet be defined; but fragmentation and rotation of crustal blocks by oroclinal folding and strike-slip faulting can be suggested. The Ellsworth Mountains, for example, consist of Cambrian-Permian metasedimentary rocks that strike northward toward the noncorrelative and latitudinally striking Mesozoic terrane of the Antarctic Peninsula in one direction and southward toward that of the Lower Paleozoic: terrane of the Transantarctic Mountains in the other; the three regions may be separated by great strike-slip faults. The Antarctic Peninsula in west Antarctica, south of South America, consists of metavolcanic and metasedimentary rocks intruded by Late Cretaceous quartz diorite. The pre-granitic rocks are of Jurassic and Early Cretaceous ages wherever they have been dated by fossils, although some crystalline complexes may be older. The S-shape of the peninsula may represent oroclinal bending within Cenozoic time as part of a motion system in which a narrow continental bridge between South America and Antarctica was deformed and ruptured. Perhaps this bridge lagged behind as the larger continental plates drifted into the Pacific Ocean Basin.

On potassium-argon age measurements of basaltic rocks from Southern Brazil

Twelve potassium-argon age measurements have been carried out on whole-rock samples and several feldspar concentrates from different samples of a diabase sheet that crops out in Laranjal Paulista, SP, Brazil. The ages obtained average 128 m.y. and the variation observed is of the same order or lower than the possible analytical error. The measurements made on whole-rock samples, on different feldspar concentrates, and on separate plagioclases coincide precisely and therefore the whole-rock procedure is regarded as suitable for use in routine working for dating purpose of the basaltic rocks from Southern Brazil. Moreover, the agreement of ages suggests that the peculiar alteration usually observed throughout these rocks is probably deuteric essentially contemporaneous with their formation.

Geology of the Central Portion of the Queen Maud Range, Transantarctic Mountains

The geologic section consists of a folded and metamorphosed basement complex of geosynclinal and nearshore sediments and intrusives, a thick sequence of nearshore and terrestrial sediments of middle to late Paleozoic age, and thick diabase sheets and basalt flows of Jurassic age. Block faulting, probably through during the Miocene age, produced the which has been carved into its present from by glaciers.

Late Newark fault versus pre-Newark peneplain in Connecticut

Inferred structural features in S.-central Connecticut previously presented in favor of a pre-Newark peneplain have not been confirmed by recent detailed geologic mapping. Part of the western boundary of the E.-dipping Newark Group [Triassic] formerly interpreted as an unconformable contact now is known to be a fault of late Newark or post-Newark age. The irregularity of the boundary near Bethany Mountain is the result of differential erosion where relief on the pre-Newark surface is moderately high. There is no evidence for the oblique faults formerly inferred there to offset a pre-Newark peneplain. An irregular diabase sheet near the base of the Newark Group in that area is clearly dikelike at Bethany Gap and cannot be used as a "marker bed" for recognizing concealed faults.

Late Newark Fault Versus pre-Newark Peneplain in Connecticut

Inferred structural features in S.-central Connecticut previously presented in favor of a pre-Newark peneplain have not been confirmed by recent detailed geologic mapping. Part of the western boundary of the E.-dipping Newark Group [Triassic] formerly interpreted as an unconformable contact now is known to be a fault of late Newark or post-Newark age. The irregularity of the boundary near Bethany Mountain is the result of differential erosion where relief on the pre-Newark surface is moderately high. There is no evidence for the oblique faults formerly inferred there to offset a pre-Newark peneplain. An irregular diabase sheet near the base of the Newark Group in that area is clearly dikelike at Bethany Gap and cannot be used as a marker bed for recognizing concealed faults.

Cobalt-arsenic minerals associated with aplites, at Cobalt, Ontario

In the Bass Lake area, about 3 miles from the town of Cobalt, soda-rich aplites are developed in the upper part of the Nipissing diabase sheet. Associated with these aplites are the following metallic minerals: pyrite, skutterudite, glaucodot, gersdorfite, rammelsbergite, chalcopyrite, tenantite, niccolite, bismuthenite. The aplites, with the associated metallic minerals, are thought to be late-stage differentiates from the tholeiitic magma from which the diabase was formed and indicate that such a magma may yield minerals carrying Co, As, S, Ni, Ag, and Bi. The major mineralization of the Cobalt region is ascribed to a similar magma in depth.

PETROLOGY OF GRANOPHYRE IN DIABASE NEAR DILLSBURG, PENNSYLVANIA

Small bodies of granophyre occur in the upper part of diabase bodies of Triassic age in southeastern Pennsylvania. One near Harrisburg was penetrated by a diamond-drill. Drill core specimens show a gradation from diabase to granophyre. New data include 10 chemical analyses, spectrographic determinations of trace elements, and the results of petrographic study of specimens from the drill core. The sequence, from diabase to granophyre, includes a chilled zone that represents an original magma of tholeiitic composition, normal diabase, pegmatitic facies of diabase, and granophyric diabase that is intermediate in composition and petrographic characteristics between diabase and granophyre, and finally granophyre. Alkalies and silica increase progressively from diabase to granophyre; iron increases to a maximum in transitional granophyric diabase, then decreases in the granophyre. It is concluded that crystal fractionation in a large sheetlike body of tholeiitic magma yielded a small amount of granophyre. Prior to complete solidification, a residual liquid rich in iron, alkalies, and silica accumulated locally in the upper part of the diabase sheet. In places volatile-rich iron-bearing solutions escaped into the overlying sedimentary rocks and deposited magnetite; the remaining liquid crystallized t o fine-grained granophyre.