Of the 21 samples acquired for the Viking X ray fluorescence spectrometer, 17 were analyzed to high precision. Compared to typical terrestrial continental soils and lunar mare fines, the Martian fines are lower in AI, higher in Fe, and much higher in S and CI concentrations. Protected fines at the two lander sites are almost indistinguishable, but concentration of the element S is somewhat higher at Utopia. Duficrust fragments, successfully acquired only at the Chryse site, invariably contained about 50% higher S than fines. No elements correlate positively with S, except CI and possibly Mg. A sympathetic variation is found among the triad Si, AI, Ca; positive correlation occurs between Ti and Fe. Sample vafiabilities are as great within a few meters as between lander locations (4500 km apart), implying the existence of a universal Martian regolith component of constant average composition. The nature of the source materials for the regolith fines must be mafic to ultramafic. INTRODUCTIONFor the first in situ analysis of the inorganic chemical composition of surface material on the planet Mars, miniature energy-dispersive X ray florescence spectrometers were designed to fit within an available space on the Viking lander. Spacecraft constraints upon configuration, deployment, mass, and heat sterilizability of the experiment package limited the instrument design. It was hoped that three to five Martian samples could be analyzed for about a dozen elements. Altogether 21 samples were delivered to the two instruments, and 15 elements were analyzed for in most samples. Descriptions of the instrument design have been published previously Toulmin et al., 1973;Clark et al., 1977]. Preliminary reports on analytical results at both landing sites were published soon after the initial data were received Baird et al., 1976]. In this paper, we present our findings for the major and minor element concentrations in all samples taken at the two landing sites. These are based upon extensive data correction procedures and upon laboratory simulations using a flight-qualified instrument identical to those on Mars, operated under Martian conditions of temperature, pressure, and atmospheric composition.

The elemental analyses whose basis is described in the preceding two papers represent the composition of samples of Martian fines; the only undetermined major constituents thought to be present are H2O, CO2, Na2O, and possibly NOx. The samples are principally silicate particles, with some admixture of oxide and probably carbonate minerals; the fines appear to have been indurated to a variable degree by a sulfate‐rich intergranular cement. The overall elemental composition is dissimilar to any single known mineral or rock type and apparently represents a mixture of materials. Close chemical similarity among samples at each site, and between the two sites, indicates effective homogenization of the fines, presumably by planetary windstorms, and further suggests that the samples analyzed represent the fine, mobilizable materials over a large part of the planet's surface. Low trace element, alkali, and alumina contents suggest that the great preponderance of the materials in the mixture is of mafic derivation; highly differentiated, salic igneous rocks or their weathering products are insignificant components of the samples. Normative calculations, comparisons with reference libraries of analytical data, and mathematical mixture modeling have led to a qualitative mineralogical model in which the fines consist largely of iron‐rich smectites (or their degradation products), carbonates, iron oxides, probably in part maghemite, and sulfate minerals concentrated in a surface duricrust. The original smectites may have formed by interaction of mafic magma and subsurface ice, and the sulfates (and carbonates?) may have been concentrated in the surface crust by subsurface leaching, upward transport, and evaporation of intergranular moisture films. Testing and refinement of this and competing models will accompany continuing acquisition of samples and data and refinement of the analyses, particularly with respect to the critical light elements Mg, Al, and Si.

No abstract

Elemental analyses of fines in the Martian regolith at two widely separated landing sites, Chryse Planitia and Utopia Planitia, produced remarkably similar results. At both sites, the uppermost regolith contains abundant Si and Fe, with significant concentrations of Mg, Al, S, Ca, and Ti. The S concentration is one to two orders of magnitude higher, and K(<0.25 percent by weight) is at least 5 times lower than the average for the earth's crust. The trace elements Sr, Y, and possibly Zr, have been detected at concentrations near or below 100 parts per million. Pebblesized fragments sampled at Chryse contain more S than the bulk fines, and are thought to be pieces of a sulfate-cemented duricrust.

It is proposed that the high sulfur measured in Martian fines may represent more than enrichment by upward migration of soluble sulfate salts. It may in fact reflect intrinsic surface abundances of S much higher than for the earth or moon. High S in the regolith is permitted by several alternative models including remnant primitive lithosphere, a primitive component in the regolith, or trapping of volcanic gases. The alternative models encompass various features of the accretion chemistry, thermal history, and core size-mass relationships that have been proposed for Mars. Possible consequences of ubiquitous sulfur include mantle mineral assemblages with unusual properties, significant sulfide content of unweathered rocks, high initial weathering rates and susceptibility to erosion of the surface, very high level of incorporation of H•_O and O•_ by surface fines, recycling of carbon and nitrogen, paleoclimatic effects of extinct atmospheric sulfur-containing gases, and a previous, if not contemporary, basis for biological activity without photosynthesis. The Martian regolith may also contain heavy volatile elements in relatively high abundance and a significant component of late accretionary material.

Chemical results from four samples of martian fines delivered to Viking landers 1 and 2 are remarkably similar in that they all have high iron; moderate magnesium, calcium, and sulfur; low aluminum; and apparently very low alkalies and trace elements. This composition is best interpreted as representing the weathering products of mafic igneous rocks. A mineralogic model, derived from computer mixing studies and laboratory analog preparations, suggests that Mars fines could be an intimate mixture of about 80 percent iron-rich clay, about 10 percent magnesium sulfate (kieserite?), about 5 percent carbonate (calcite), and about 5 percent iron oxides (hematite, magnetite, maghemite, goethite?). The mafic nature of the present fines (distributed globally) and their probable source rocks seems to preclude large-scale planetary differentiation of a terrestrial nature.

Ten samples of the Martian regolith have been analyzed by the Viking lander X ray fluorescence spectrometers. Because of high‐stability electronics, inclusion of calibration targets, and special data encoding within the instruments the quality of the analyses performed on Mars is closely equivalent to that attainable with the same instruments operated in the laboratory. Determination of absolute elemental concentrations requires gain drift adjustments, subtraction of background components, and use of a mathematical response model with adjustable parameters set by prelaunch measurements on selected rock standards. Bulk fines at both Viking landing sites are quite similar in composition, implying that a chemically and mineralogically homogeneous regolith covers much of the surface of the planet. Important differences between samples include a higher sulfur content in what appear to be duricrust fragments than in fines and a lower iron content in fines taken from beneath large rocks than those taken from unprotected surface material. Further extensive reduction of these data will allow more precise and more accurate analytical numbers to be determined and thus a more comprehensive understanding of elemental trends between samples.

Major-element analyses of 497 composite samples of batholithic rocks (quartz diorites, granodiorites, and quartz monzonites) from the northern Peninsular Ranges and Transverse Ranges Provinces, southern California, form the basis for a mixing model that accounts for most of the compositional variation in the rocks. The compositional structure in the batholithic rocks as a group was found to be similar to that in the Sierra Nevada batholith, and indicates that four end members are sufficient to account for 85-97 percent of the variability in each of the eight major oxides. According to the model, the batholithic rocks formed from the mixing of basaltic and quartzofeldspathic end-member magmas, and the removal of variable proportions of plagioclase and mafic minerals, principally hornblende. Gabbroic rocks, common only in the western part of the region, could have formed from nearly uncontaminated magmas of the basaltic end member. Variations in the mixtures of basaltic and quartzofeldspathic magmas are presumed to reflect variations in their source materials at depth. According to the model, the compositions of the source materials do not vary smoothly over the region, but display a discontinuity along a line approximately coincident with the present San Jacinto fault zone. The discontinuity is roughly coincident with previously noted petrologic and isotopic discontinuities in the northern Peninsular Ranges and is interpreted as the western limit of significant contribution of continental materials to the batholithic rocks. The model gives no evidence of a discontinuity in the vicinity of the San Andreas fault zone. 'Average of eight analyses from Larsen and Draisin (1950, p. 69). 2Analysis 7 from Larsen and Draisin (1950, p. 71). ' See figure 3 for sampling localities and Appendix 1 for the analyses.