Library

Notes: Abstract Si87/Sr86 ratios have been determined for lavas and py lastic rocks from three basalt-andesite centers along the Lesser Antilles arc—Mt. Misery on the island of St. Kitts, Soufriere on the island of St. Vincent, and Carriacou, an island of The Grenadines. The average Si87/Sr86 content of these rocks is 0.7038 for Mt. Misery, 0.7041 for Soufriere, and 0.7053 for Carriacou. All the Sr87/Sr86 values from each center are the same within analytical uncertainty (±0.0002). The constancy of strontium isotopic data within each center supports the hypothesis that basalts and andesites for each specific center investigated are generated from the same source — in agreement with petrographic and major- and minor-element data. Strontium isotopic compositions and elemental concentrations, particularly of strontium and nickel, indicate that this source was mantle peridotite and that the relationship between the respective basalts and andesites is probably fractional crystallization.

Notes: Abstract Basalts in the Southern Rocky Mountains province have been analyzed to determine if any of them are primitive. Alkali plagioclase xenocrysts armored with calcic plagioclase seem to be the best petrographic indicator of contamination. The next best indicator of contamination is quartz xenocrysts armored with clinopyroxene. On the rocks and the region studied, K2O apparently is the only major element with promise of separating primitive basalt from contaminated basalt inasmuch as it constitutes more than 1 % in all the obviously contaminated basalts. K2O: lead (〉 4 ppm) and thorium (〉 2 ppm) contents and Rb/Sr (〉 0.035) are the most indicative of the trace elements studied. Using these criteria, three basalt samples are primitive (although one contains 1.7% K2O) and are similar in traceelement contents to Hawaiian and Eastern Honshu, Japan, primitive basalts. Contamination causes lead isotope ratios, 206Pb/204Pb and 208Pb/204Pb, to become less radiogenic, but it has little or no effect on 87Sr/86Sr. We interpret the effect on lead isotopes to be due to assimilation either of lower crustal granitic rocks, which contain 5–10 times as much lead as basalt and which have been low in U/Pb and Th/Pb since Precambrian times, or of upper crustal Precambrian or Paleozoic rocks, which have lost much of their radiogenic lead because of heating prior to assimilation. The lack of definite effects on strontium isotopes may be due to the lesser strontium contents of granitic crustal rocks relative to basaltic rocks coupled with lack of a large radiogenic enrichment in the crustal rocks. Lead isotope ratios were found to be less radiogenic in plagioclase separates from an obviously contaminated basalt than in the primitive basalts. The feldspar separate that is rich in sodic plagioclase xenocrysts was found to be similar to the whole-rock composition for 206Pb/204Pb and 208Pb/204Pb whereas a more dense fraction probably enriched in more calcic plagioclase phenocrysts is more similar to the primitive basalts in lead isotope ratios. The primitive basalts have: 206Pb/204Pb ∼ 18.09–18.34, 207Pb/204Pb ∼ 15.5, 208Pb/204Pb ∼ 37.6–37.9, 87Sr/86Sr ∼ 0.704–0.705. In the primitive basalts from the Southern Rocky Mountains the values of 206Pb/204Pb are similar to values reported by others for Hawaiian and eastern Honshu basalts and abyssal basalts, whereas 208Pb/204Pb tends to be equal to or a little less radiogenic than those from the oceanic localities. 87Sr/86Sr appears to be equal to or a little greater than those of the oceanic localities. These 206Pb/204Pb and 208Pb/204Pb ratios are distinctly less radiogenic and 87Sr/86Sr values are about equal to those reported by others for volcanic islands on oceanic ridges and rises.

Notes: Abstract The strontium isotopic compositions have been determined for twelve tholeiitic basalts dredged from the Gordo and Juan de Fuca Rises. Sr87/Sr86 ratios range from 0.7012 to 0.7031 and average 0.7026. These data, combined with other data from the East Pacific Rise indicate that tholeiite basalts being erupted along the active rises, in the Pacific Ocean, contain less radiogenic Sr87 than basalts erupted on the islands. These isotopic differences between the ocean-ridge tholeiite and the more alkali island basalts indicate that variations in Rb/Sr have persisted in the mantle for billions of years. The possible origins and distribution of these heterogeneties are discussed.

Notes: Abstract Rocks of the northeast portion of the Colorado mineral belt form two petrographically, chemically and geographically distinct rock suites: (1) a silica oversaturated granodiorite suite; and (2) a silica saturated, high alkali monzonite suite. Rocks of the granodiorite suite generally have Sr contents less than 1000 ppm, subparallel REE patterns and initial 87Sr/ 86Sr ratios greater than 0.707. Rocks of the monzonite suite are restricted to the northeast part of the mineral belt, where few rocks of the granodiorite suite occur, and generally have Sr contents greater than 1000 ppm, highly variable REE patterns and 87Sr/86Sr initial ratios less than 0.706. Despite forming simple, smooth trends on major element variation diagrams, trace element data for rocks of the granodiorite suite indicate that they were not derived from a single magma. These rocks were derived from magmas having similar REE patterns, but variable Rb and Sr contents, and Rb/Sr ratios. The preferred explanation for these rocks is that they were derived by partial melting of a mixed source, which yielded pyroxene granulite or pyroxenite residues. The monzonite suite is chemically and petrographically more complex than the granodiorite suite. It is subdivided here into alkalic and mafic monzonites, and quartz syenites, based on the textural relations of their ferromagnesian phases and quartz. The geochemistry of these three rock types require derivation from separate and chemically distinct magma types. The preferred explanation for the alkalic monzonites is derivation from a heterogeneous mafic source, leaving a residue dominated by garnet and clinopyroxene. Early crystallization of sphene from these magmas was responsible for the severe depletion of the REE observed in the residual magmas. The lower Sr content and higher Rb/Sr ratios of the mafic monzonites requires a plagioclase-bearing source. The Sr-isotope systematics of the majority of these rocks are interpreted to be largely primary, and not the result of crustal contamination. The positive correlation of Rb/Sr and 87Sr/86Sr ratios for the least fractionated samples indicate that the sources from which parent magmas of both the granodiorite and monzonite suites were derived are Precambrian in age.

Notes: Abstract Basalts dredged from the Mid-Indian Ocean Ridge System have rare earth, Rb, and Sr concentrations like those from other mid-ocean ridges, but have slightly higher Sr87/Sr86 ratios. Underlying gabbroic complexes are similar to the basalts in Sr87/Sr86, but are poorer K, Rb, and in rare earths. The chemical and isotopic data, as well as the geologic relations suggest a cumulate origin for the bulk of the gabbroic complexes.

Notes: Abstract Phenocrysts of sodic sanidine from twelve upper Cenozoic units of silicic ash-flow tuff and lava from the Western United States contain from 0.25 to 0.45 the Rb present in the associated groundmass materials. The ratios of potassium to rubidium in the sanidines are, on the average, about four times greater than those of the groundmass. Separation of phenocrystic sanidine from salic melts provides an efficient method for raising the Rb content and lowering the K/Rb ratio of the melts, although the amount of differentiation probably is limited by continuous reequilibration of the alkalis between crystal and liquid phases through ion exchange. Syenites of cumulate origin will have appreciably lower Rb contents and higher K/Rb ratios than the melts from which they precipitated. Available data on the distribution of Rb between synthetic biotite and K-sanidine demonstrate that the separation of biotite probably will not deplete salic melts in Rb relative to K.

Notes: Abstract Concentrations of Rb, Sr, and REE (rare earth elements), and Sr-isotopic ratios in rocks of the Cambrian alkaline complexes in the Wet Mountains area, Colorado, show that rocks formed as end-products of a variety of magmas generated from different source materials. The complexes generally contain a bimodal suite of cumulus mafic-ultramafic rocks and younger leucocratic rocks that include nepheline syenite and hornblende-biotite syenite in the McClure Mountain Complex, nepheline syenite pegmatite in the Gem Park Complex, and quartz syenite in the complex at Democrat Creek. The nepheline syenite and hornblende-biotite syenite at McClure Mountain (535±5m.y.) are older than the syenitic rocks at Democrat Creek (511±8m.y.). REE concentrations indicate that the nepheline syenite at McClure Mountain cannot be derived from the hornblende-biotite syenite, which it intrudes, or from the associated mafic-ultramafic rocks. REE also indicate that mafic-ultramafic rocks at McClure Mountain have a source distinct from that of the mafic-ultramafic rocks at Democrat Creek. In the McClure Mountain Complex, initial87Sr/86Sr ratios for mafic-ultramafic rocks (0.7046±0.0002) are similar to those of hornblende-biotite syenite (0.7045±0.0002), suggesting a similar magmatic source, whereas ratios for carbonatites (0.7038±0.0002) are similar to those of nepheline syenite (0.7038±0.0002). At Democrat Creek, initial ratios of syenitic rocks (0.7032±0.0002) and mafic-ultramafic rocks (0.7028±0.0002) are different from those of corresponding rocks at McClure Mountain.

Notes: Abstract We present initial 176Hf/177 Hf ratios for many samples of continental crust 3.7-0.3 Gy old. Results are based chiefly on zircons (1% Hf) and whole rocks: zircons are shown to be reliable carriers of essentially the initial Hf itself when properly chosen on the basis of U-Pb studies. Pre-3.0 Gy gneisses were apparently derived from an unfractionated mantle, but both depleted and undepleted mantle are evident as magma sources from 2.9 Gy to present. This mantle was sampled mainly from major crustal growth episodes 2.8, 1.8 and 0.7 Gy ago, all of which show gross heterogeneity of 176Hf/177Hf in magma sources from εHf=0 to +14, or about 60% of the variability of the present mantle. The approximate εHf=2εNd relationship in ancient and modern igneous rocks shows that 176Lu/177Hf fractionates in general twice as much as 147Sm/144Nd in mantle melting processes. This allows an estimation of the relative value of the unknown bulk solid/liquid distribution coefficient for Hf. DLu/DHf=∼ 2.3 holds for most mantle source regions. For garnet to be an important residual mantle phase, it must hold Hf strongly in order to preserve Hf-Nd isotopic relationships. The ancient Hf initials are consistent with only a small proportion of recycled older cratons in new continental crust, and with quasi-continuous, episodic growth of the continental crust with time. However, recycling of crust less than 150 My old cannot realistically be detected using Hf initials. The mantle shows clearly the general positive εHf resulting from a residual geochemical state at least back to 2.9 Gy ago, and seems to have repeatedly possessed a similar degree of heterogeneity, rather than a continuously-developing depletion. This is consistent with a complex dynamic disequilibrium model for the creation, maintenance and destruction of heterogeneity in the mantle.