of...sea sediments. “ igneous’’ rocks under this heading we include some ultra- basic rocks...

KARL K. TUREKIAN Dept. Geology, Yale Uniucrsity, New Haven, Conn. KARL HANS WEDEPOHL Mineralogische-Instilut der Uniuedat, Gottingen, Germany

Distribution of the Elements in Some Major Units of the Earths Crust

Abstract: This paper presents a table of abundances of the elements in the various major units of the Earth’s lithic crust with a documentation of the sources and a discussion of the choice of units and data.

CONTENTS

Introduction . . . . . . . . . . . . . . . . 175 References cited . . . . . . . . . . . . . . . 187 Choice of units . . . . . . . . . . . . . . . 175

General statement . . . . . . . . . . . . . 175 “Igneous rocks” . . . . . . . . . . . . . . 175 1. Estimation of hafnium concentrations in Sedimentary rocks . . . . . . . . . . . . . . 176 “igneous” rocks . . . . . . . . . . . . 183 Deep-sea sediments . . . . . . . . . . . . . 176 2. Distribution of the elements in the Earth’s crust.

Clioicc of data . . . . . . . . . . . . . . . . 177 Metamorphic rock; . . . . . . . . . . . . 177 . . . . . . . . . . . . . . . Pacing 18G

INTRODUCTION Several tables of the crustal abundances of

the elements have been published to date (Rankama and Sahama, 1950; Goldschmidt, 1954; Flcischcr, 1353; Vinogradov, 1956; Riason. 19%) either as parts of treatises on thc geochemistry of thc elements or as attcrnpts to compile a list for general use. In addition, Green (1959) and Vinogradov (1566) have published charts of the distribution of many elements in various units of the Earth’s crust.

We have found these tablcs deficient in some aspects. This awareness arose when the two of us independently were preparing articles on the geochemical distribution of the elements for the Encyclopedia of Science and TechnoloKy published by hicGraw-Hill (Turekian, 1960) and for the new edition of Lehrbuch der Geologie, Ted I . by E. Kayser and R. Brink- mann, to be published by F. Enke, Stuttgart (Wedepohl).

The individual tables in these two works have been modified and collated here (Table 2) with a fuller description of the plan used in compiling the data since a brief summary article of the sort required for the encyclo- pedias offered no possibility of presenting the Sources of inlormation wed.

Any compilation is neccssarily subject to great uncertaintics in thc reliiibility o l thc analytical work, the sampling, and the in- tcrpretations, both of the original iiivcstigator and the compilcr. Hence the accompanying table should bc accepted not so rnuch as a doctrine but as a motion on the floor to be dc- batcd, and amcndcd or rejected.

CHOICE OF EXITS General Statement

With the wide diversity of rock types aviiil- able for sampling in the Earth’s crust thc choice of units for a compilrition must to some degree be arbitrary. \\:e have chosen three major groups for the presentation of the data, “ igneous” rocks, sedimentar> rocks. and deep- sea sediments.

Igneous’’ Rocks “

Under this heading we include some ultrabasic rocks and all basaltic rocks as being of undoubted igneous origin. Granitic and syenitic rocks, even though they d o not all show unequivocal evidence for igneous origin, are included under “igneous” rocks for the d e of simplicity.

Peridotitic rocks \\’ere chosen whcnever pos-

Geologicnl Socicty of :\inrrica Bulletin, v. 72, p. 175-192, February 1961

175

. . . . - . . . . . .

. .

. . . -176 . TURERIAN AND WEDEPOHL-DISTRIBUTION OF ELEMENTS

sible to represent the ultrabasic group. Ultra- basic rocks with unusual metamorphic histories wcre generally avoided in compiling the trace- element data. Serpentines were also avoided because several elements (boron, arsenic, and germanium, etc.) are notably enriched in these altered rocks relative to dunites and peridotites.

The basaltic rocks include all manifestations of rocks of basaltic composition, i.e., gabbros, dolerites, and basalts. Concentrations of some elements show differences between the intrusive and the extrusive or hypabyssal representa- tives. In only a few cases were the differences significant in terms of the information available. In those cases the extrusive and hypabyssal rocks were weighted more heavily than tlie intrusive rocks to arrive ? t the figure in Table 2.

Tlie granitic rocks afford some difliculty in classification. ;\I1 rocks associatccl with ;I granitic terrane are considered granitic rocks a1- though local or widcr variations yield n variety of rock types such as granodioritc, q u x i z monzonite, etc. With such a widc variety of possibilities in rock types and chc vngnrics of meaning of somc of the noincnclaturc prc- scntcd in the litcraturc, we decided that two categories of granitic rocks were all that wcrc practical for the present compilation. Uiicicr tlie bias of a previous such considcration ncccssary i n evaluating the geochemistry of strontium (Turckian and Kulp, 1956). we have chosen the groups in terms of their expected calcium concentrations. uiz., high-calcium granitic rocks with a mean gross chcmical composition of a granodiorite, and low-calcium granitic rocks with a composition approaching that ofan ideal granite. This choice is arbitrary. Tlie prcscnted data are just not any better than this gross clnssification.

.kcording to the esperience of field geologists, granites. granodiorites, and basaltic rocks are by far the most common rock types. We include the syenites as a type in spite of their subordinate abundance. We have generally tried to weigtit the values toward the syenite rather than the nepheline syenite end because the latter type is the rarer.

In the case of the granitic and syenitic rocks w e have avoidcd using their extrusive equiva- lents in computing the averages. For several trace elements the extrusive acidic rocks are very diflcrent in their abundance from thc intrusive chemical equivalent. Rhyolites have variable and perhaps unusual afini ties.

Sedimentary Rocks .The standard breakdown of sedimentary

rocks is into shales, sandstones, and carbonate rocks as end members, and other rocks as mixtures of these. This classification is based on sequences associated with Kay’s (1951) miogeosynclinal areas where reasonably thor- ough chemical degradation of the original source rock is supposed to have occurred. There are, of course, vast amounts of sedimentary rocks which are composed to a large degree of poorly sorted more or less degraded minerals, uiz.. conglomerates, arkoses, and graywackes. These rocks represent a great problem in the presentation of data on scdiinentary rocks. I t is not possible to dismiss them as the mechani- cal degradation products of weatlicring and sedimentation sincc in these processes a chemical diflircnti;ition from the original rocks must havc taken place. [-Io\vevcr, because of the great complexity of thesc rock types and an unccrtainty ns to their nianncr of origin they are not iiicliidctl iri tlie accompanying table. It must be noted th;rt this is an omission bccausc of lack of iiiforniation rather than becausc of unimportrincc. Macplicrson (1958) reports that Can:idi:in Prccambrian argillitcs and lo\v-gr:idc schists Iinvc the samc composition for many tracc clcincnts :IS ;issociatcd gray\vackcs. In ;iJdition. \\:cbcr‘s (1360) data seem to indicate that n tvidc rangc of gray- ivackcs have similar composition with regard to most of tlie tracc elcments (zirconium seems to be an csccption).

Deep-sea Sediiiientj

Deep-sea sediments c ~ i i n o t rightly be classi- fied under the term “rock“ since much of the sampling is done on materiiil which exists permeated continuously by sen water and has not yet been subjected to lithification or ex- treme diagenesis.

Two end membcrs only are considered: the pelagic clay, essentially free of calcium carbonate; and the carbonate-rich sediment in its purest sampled form containing about I O per cent clay fraction. Further. following Goldberg and Arrhenius (195s). we assume that the dis- solved solids i n thc n’ater permeating the sediment are part of tlie sediment rather than of the hydrosphcre. This incans that analyses on unwashcd samples are preferred. Estimating the abundance of several of the elements in the deep-sea matcrial is complicated bv the fact

that their concentration is greater in the Pacific sediments than in the Atlantic. Al- though the Pacific is roughly three times larger than the Atlantic in area, the rate of sedimentation may be about three times greater in the Atlantic basin (Wedepohl, 1960) than in the Pacific. This being the case, where the above disparity is observed, a simple average of the Atlantic and Pacific values was used for the abundance table.

Metamorphic R O C ~ J We have assumed that metamorphic rocks

generally retain a chemical composition similar to their unmetamorphosed equivalent. How- ever, often a schist is sampled free of quartzo- feldspathic segregations. In such a case the schist will be higher than tlie original rock in tlie concentrations of thc elcmcnts associated with the mafic minerals. The \\hole rock, however, will probably show the composition of the original unmetamorphosed rock. Where metamorphism grades into granitization, the granitized rock is placed in the chemical catcgory of granitic rocks, hence not treated sepxa tcl y.

CHOICE OF D.lT.4 Generally the ne\vcst information \ v x used

to construct the table ivhcncvcr avail:iblc. hluch iicw work has been Jonc on tlic trace elements since the end of \Vorld \\’:ir I 1 and particulxly since 1950.

The following is an element-by-elcment discussion of the sources of the information of Table 2 . We have delibcratcly used the first person in writing becausc the table represents solely our judgement in compilation. There is always the risk that when such a table is published the sources and uncertainties in i t may be forgotten and “the table” qiiotcd un- critically. This must be avoided.

Lithirim: Thedata are primarily from Horst- man (1957). HoIvever, his ultrabasic value of 26 ppm is not used since it is considerably higher than that of Strock (1936), who gives 2 ppm, and that of Pinson, .-\hrens, and Franck (1953), who give <0.3 ppm. The value for carbonates is an upper limir. and tlie value of carbonate dcep-sea corcs is based on the assumption that even the purest pelagic calcareous cores have approsimatcly I O per cent clay fraction which contributes the lithium.

The data for granitic and biisalric rocks are taken from Sandcll (1952). h*icrrill,

BerjtIIiwn:

. . Honda, and Arnold (195s) report 3.3 ppm for G-1 standard granite and 0.68 ppm for W-1 standard diabase. The nepheline syenite value is from Borodin (1956). This is lower than the concentrations given by Goldschmidt (1954) and Holser et ai. (1951). Merrill et ai.

’ (1960) have analyzed four pelagic clay cores from the Pacific and one from the Atlantic and find very small variations in the beryllium concentration. Their average of 2.6 pprn is used here. Other data appearing in the literature for pelagic clays range from 1.1 ppm (Goel et al., 1957) to 8 pprn (Tatsumoto, 1957). Since tlie beryllium concentrations of the Atlantic and Pacific pelagic sediments are not different, although the Atlantic and Pacific have ditferent accumulation rates, we assume that the beryllium is closely associated with the.clay minerals. Hence we have assumed thar shales will have the same composition :is pelagic clays rathcr than the 6 ppm reported by Goldschmidt (1954).

Goldschmidt’s data sccm high for this clc- ment in all rock typcs comparcd to tlic current data. Tlie sandstone, carbonate, and carbonatc dccp-sea sediment data arc lacking, but prolxi- bly the abuntlance in cach rock typc is of tlic ordcr of tenths of parts pcr million.

Tlie ultrabasic. basdtic. granitic. a n d sycnitic values :irc from l-larder (1959a; 195%). The granitic rocks prcsent sonic problems of intcrprctation. S;iliaina’s (See Rankaina and Sahaina, 1950) low values (3-10 pprn) for Fcnnoscandian rocks may be compared to Wasserstein’s (1951) values for somc South African granites which run up to 150 ppm. Okada (1955; 1956) found boron conccntra- tions in Japanese granitic rocks ranging from 1 to 160 ppm. Since boron is a highly mobile element . . during metamorphic and igneous activity, the wide range of values may be expected. Granitic rocks from roof areas of intrusive rocks and granitic migmatites generally have higher concentrations of boron. Degens, Williams, and Keith (1957) and Harder give an average value for shales of 100 ppm. The deep-sea clay value is the averagc of the boron content of Pacific clays (Goldberg and Arrhenius, 1955) and .-\tlantic clays analyzed by Harder (1959b). Tlie carbonate deep-sea sediment value is based on a few analyses of Atlantic material made by Harder, and tlic sandstone and carbonate rock values are :ilso his.

Tlicre arc nvo current sets of de-

.

Iloron:

Ni/rogen:

178 TUREKIAN AND WEDEPOHL-DISTRIBUTION OF ELEMENTS

terminations of the nitrogen content of some igneous rocks. F. Wlotzka (1960, Ph.D. thesis, Gottingen Univ.) reports about 30 ppm for basaltic rocks and 20 ppm for granitic rocks. R. S. Scalan (1959, Ph.D. thesis, Univ. of Arkan- sas), in studying the isotope geochemistry of nitrogen, determined the composition of a number of ultrabasic (average 6 ppm) and basaltic (average 17 ppm) rocks. Because of the wide range of values within each rock type we have chosen a value of 20 pprn for each igneous rock except for ultrabasic rocks, for which Scalan's average of 6 pprn is used, and syenites, for which Wlotska's average of 30 pprn N is used. Both investigators indicate that the main form of the nitrogen is as the NHl+ ion.

Since sediments have greatly variable nitrogen concentrations, mainly a function of the organic content of the sediment, these estimates are not included in the tablc.

Fluorine: I\.lost of the values in Tiiblc 2 are from Korirnig (1951) as mociilicd in thc following cascs by information from 0 t h workcrs. R. H. Seraphim (1351, P1i.D. thcsis, hlass. Inst. of Technology) and Kokubu (1956) list 520 pprn and 530 ppm respcctivcly for granitic rocks they analyzed; this agrccs closely with Koritnig's i ~ ~ l u c . On the othcr hand I<okubu (most of whose rocks were Japancse) found a lorn value of 250 ppm for basaltic rocks, whereas Seraphim reportcd a value of 540 ppm, which is higlicr than Koritnig's. Korit- nig's value of 520 ppm for granodioritic rocks is used, although the average of alkali granitic and dioritic values of other authors leads to a higher value. T h e syenite valuc is the average of Koritnig's (950 ppm) and Scraphim's (1450 ppm) values. T h e carbonate value is the average of Koritnig's limestone and dolomitc analyses. Kokubu reports considerably lower values for limestones (100 pprn). The values for deep-sea sediments are taken from Sera- phim. They are based only on Atlantic Occan sediments. Shepherd's (1940) figures on sediments from the Pacific (clay, 660 ppm) seem too low.

Sodirrm: T h e igneous-rock data exccpt basalt arc from Nockolds (1954), using the nver- ages for alkali granite (his Tablc I , column III), granodiorite (Table 2 , column III), peridotite (Table 9, column I), and alkali syenite (Table 3, column IV). T h e basaltic value is an average of Green and Poldervaart's (1955) compiled mean tholeiitic and mean olivine basaltic rock. The sedimentary-rock data are Clarke's (1924). The deep-sea scdimcnts providc some difficulty

. .

since all the cores are rich in sodium chloride derived from interstitial sea water. Goldberg and Arrhenius (1958) present compelling reasons for accepting the bulk composition of the core, including the interstitial salts, as representative of the sediment, and this is done in the table. T h e data for pelagic clays are from Goldberg and Arrhenius (1958). The data for the carbonate deep-sea cores are more difficult to obtain. Broecker, Turekian, and Heezen (1958) report an average of 5 per cent NaCl in dry, unleached carbonate core material. This corresponds to a sodium concentration of around 20,000 ppm and a chlorine concentration of 30,000 ppm. T h e highest- carbonate core reported by Goldberg and .\rrhenius (1955) has 16,000 ppm Na.

Mugncsiirm: Igneous-rock data are from Nockolds (1954) and Green and Poldevaart (1955), as above. Sedimentary-rock data are from Clarke (1924). The pelagic-clay value is from Clarkc (1924) and Goldberg and Ar- rhenius ( I 958). Carbonntc dcep-sea-corc data arc from P. J. Waiigcrsky (1958, P1i.D. thesis, Y a k Univ.) and Turckian 2nd Fccly (1956), nho agrcc vcry well for :Illantic Equatorial cores.

: lhninir tv crnd silicon: Igneous-rock data arc from Nockolds (1954) and Green and Polclevaart (1955) as above: sedimentary-rock daia from Clarkc (1924); pclagic clay from Goldberg :iiid :\rrliciiius (1355); and carbonate dcep-sea-corc data from the analysis of Atlantic Equatorial Corc AISO-74 by P. J. Wangersky (1955, P1i.D. thesis, Yale Univ.).

T h c igneous-rock data are from Xockolds (1954) and Green and Poldevaart (1955) as described above. The midstone and carbonate-rock data are from Koritnig (1951). The shale value is the average previously reported by Wcdepohl (1960).

Corrcns (1937) reports 1500 ppm for Atlantic pelagic clays, which is probably a minimum for these sediments. He also found that clay-free calcareous sediments from the Atlantic had about 350 ppm phosphorus. We use his values for deep-sea sediments.

Because of the various possible forms of sulfur incorporation in geological materials, i t is dificult to assess the significance of the various data reportcd in the literature on this elemcnt. The earlicst paper giving 3 large amount of dara on sulfur in igneous rocks-is by Troger (1934) and is that used in the compilation by Rnnkama and Sahama (1950) and others. Sandcll and Goldich (1943) report three

PhospltorrrJ:

SirlJirr:

179. . .

.. . CHOICE’OF‘DATA ’ ’. . . . -

. . . . -

. .

values for sulfur in Minnesota rocks (“granite,” “diorite,” and “diabase”), which have a range of 200 to 400 pprn with no obvious relationship to rock type, and two diabases from New England with a value of 1200 ppm. Ricke (1960, in press) reports about 270 ppm S for granites and 250 pprn S for basalts. H e found that olivine contains about 30 pprn sulfur, but since ultrabasic rocks have a variable sulfide component this number cannot be used with certainty. Ricke also reports 400 pprn S for granodioritic rocks and 440 pprn S for syenites. Although one of us (Wedepohl) feels that these numbers represent the abundance of sulfur in the various igneous-rock types, we have assigned a common value for all “igneous” rock types, v t . 300 ppm S, because it may be that the variations within the rock types in ‘a wider sampling would esceed that between units. One of us (Turckian) believes that this value t$cd on Ricke’s data probably 113s only order of magnitude reliability, but even thin it is lower than some of Troger’s values.

Tlic information on scdiments anti sedimentary rocks is in no bctter shapc. Clarke (1924) reports an average value of 2600 ppm for shales, whereas Rickc (in press) gets an average of 2200 pprn for his snmpling. Higher valucs liavc becn reported by Tourtclot (1957) for thc I’icrrc shale (5500 ppm) and by othcr authors (including .\Iinami. 193%; Vinogradov and Ronov, 1956) for carbonaceous shales. Wc choose tlic average of Clarke’s and Ricke’s values for all thc sedimentary-rock types. Infor- mation on the sulfur concentration of deep-sea sediments is from Edgington and Byers (1942). Ricke got csscntially the same value for pclagic clays. Tlie sulfur concentration is obviously that of sea-salt contribution to the sediment.

Chlorine: Correns ( I 956) has recently compiled the available data on the halogens.

The ultrabasic value is the average of the anhydrous dunite analysis by Kuroda and Sandell (1954). These authors give a wide range of values for igneous rocks with averages all about 200 ppm for the various rock types, except syenitic rocks. We have used the data of Behne (1953), however, for all the rock types except the ultrabasic and syenitic. Tlie deep- sea-sediment data are contingent on the argu- ments presented under sodium. However, the clay fraction probably has somc sodium in ex- ccss of the stoichiometric amount necessary to balance the chloride. Behne reports 21,000 pprn chlorine for pelagic clays. Tlie same value is assumed for the carbonate sediments.

I

..

. . Potassium: We have chosen the low average

value of Holyk and Ahrens (1953) for ultra- . basic rocks. Nockolds (1954) gives an average of about 2000 ppm K, but this may either include mica or feldspar-rich ultrabasic rocks such as kimberlite or include analyses errone- ously high in potassium. The remaining igneous-rock data are from Nockolds (1954) and Green and Poldervaart (1955) ; sedimentary- rock data are from Clarke (1924). The pelagic- clay value is from Goldberg and Arrhenius (1958). T h e deep-sea carbonate value is based on the assumption that the potassium is in the 10 per cent clay fraction plus about 400 pprn K in the soluble salts.

Calcium: Igneous-rock data are from Nock- olds (1954) and Green and Poldervaart (1955); sedimentary-rock data from Clarke (1924), whose sandstone average may be high; pelagic- clay data from Goldberg and Arrhenius (1958); carbonate deep-sea-sediment data from Tu- rekian and Feely (1956) anti P. J. Wangersky (1955, P1i.D. thesis, Yale Univ.) who liavc similar rcsults on Atlantic Equatorial cores.

Tlie ultrabasic valuc is from Pin- son, Ahrcns, and Franck (1953). Both Nockolds and Alcn (1956) x i d :\hrens (1954) rcport mean values in basaltic rocks of about 30 ppm. T h e data for lowcalcium and high-calcium granitic rocks arc from Ahrcns (1954). [-IC reports a value of 1 I pprn for granites. If thcsc granites can be rcgnrclcd as a onc-to-onc misturc of low- calcium and high-calcium granitic rocks, as seems likely, and if tlic Sc is grcatcr by 3 factor of two in the more calcic granitic rocks than in the low-calcium granitic rocks, then low- calcium granitic rocks have 7 ppm and liigh- calcium granitic rocks have 14 ppm Sc; Sahama (1945) reports I ppm Sc for Finnish granites and Hiigi (1956) 12 pprn Sc for Swiss granites. The Sc content of syenitic rocks given is that of Sahama (1945) for Fi,nnish rocks which compares with the analysis of an Arkansas nepheline syenite (Gordon and hlurata. 1952). The shale value is from \Vedepohl (1960); it compares with that of Shaw (1954) for the Littleton formation primarily. Tlie pelagic- clay value is the average of Pacific and Atlantic values from Goldberg and ..\rrhenius (1955) and Wedepohl (1960) respectively. The carbonate deep-sea-core value is based on a 10 per cent red-clay fraction contributing the Sc. Tlic value for sandstones is Sahama’s (1945) quartzite average, and that for limestones is guessed at, assuming that they contain approsimately 10 per cent clay.

”

Srundiwn:

. -

Titanium: T h e igneous-rock data are from Nockolds (1954) and Green and Poldervaart (1955). Sandstone and carbonate-rock data are from Clarke (1924). The shale value is the average used by Wedepohl (1960). The pelagic- clay value is the average of the figures given by Wedepohl (1960) and Goldberg and Arr- henius (1958), and the deep-sea carbonate sediment value is from P. J. Wangersky (1958, Ph.D. thesis, Yale Univ.).

The ultrabasic value is derived from the data of Ross, Foster, and Myers (1954), who list values for vanadium in separated minerals from ultrabasic rocks. Using a ratio of 60 per cent olivine, 20 per cent enstatite, I O per cent chrome diopside, and 10 per cent plagioclase we arrive a t 40 ppm V. T h e value for basaltic rocks is from an average of all basaltic rocks (72) a d y z e d by Nockolds and :\llcn (1956). This figure corrcsponds with a one-to-one average of tholcii tic basalts (330 ppni V) and olivinc basalts ( I40 ppm V) from unpiiblishcd rcsults of Wcdcpolil. T h c gmii t ic v;ilues are derived from ;\hrcns (1954), using the s ime assumptions as those usctl to clcrivc the scandium numbers. I-Iiigi (1955) gcts SO ppm V for granitic rocks of thc :\arc-massiv. For syenitic rocks Siiliama (1945) rcports 30 ppin V; Butler (1954) got lcss t l im IO ppm in one sample. Gordon mid hlurata (1352) list a value of 47 ppni for an :\rkansas ncplicline sycnitc. We choose 30 ppm lor this rock type.

'rllc shale value is from \Vcdcpohl (1960); i t corresponds with Jost's (1932) and Sl~aw's (1954). Degcns. Williams, and Keith (1957) rcport ;I l o w r value for Carbonifcrous shalcs of Pcnnsylvania (44 ppm). The data for sandstones and limestones are from Goldschmidt (1954), who quotes the data of Jost primarily. The limestone value agrees with the average of eight British limestones analyzed by Hirst and Xicholls (19%). T h e pelagic-clay value is from Goldberg and .-Irrhenius (1958) and Wedepohl (1960). The carbonate deep-sea-sediment value lies between 1 and 3 ppm V (Wcdepohl, 1955); hence we choose an average of 2 ppm.

T h e ultrabasic value for chromium is derived from the data in ROSS, Foster. and Myers (195-1). These authors give chromium values for separated minerals from ultrabasic rocks. There is not much variation for any one mineral type. \Ve again use the arbitrarily de- fined ultrabasic rock of the following mincr- alogic composition: 60 per cent olivine, 20 per cent enstatite, 10 per cent chrome diopside, and I O per cent plagioclase. T h e resulting value

Vanadium:

Chromizrm:

I

180 TUREKIAN AND WEDEPOHL-DISTRIBUTION OF ELEiMENTS

of 1600 pprn Cr may be too low if chromite is a very important accessory. A single analysis of a dunite by activation analysis reported by Turekian and Carr (1960), however, confirms this low value. The basalt value is from Turekian (1956). Frohlich (in press) reports 70 ppm for tholeiitic basalt and 280 ppm for olivine basalt. T h e low-calcium granitic value is from Turekian and Carr (1960) based on neutron activation analyzed rocks. T h e high- calcium value has been changed from our previous 27 ppm Cr reported in the paper just cited to 22 ppm C r as the result of additional work to be published soon. These numbers are lower than those of Ahrens (1954). The syenitic rock value is from Gordon and Murata (1952) and Butler (1954). The shale value is an average of the data of Sliaw (1954), Frohlich (in press), and Turekian (unpublished). Frohlich reports an avcragc of 15 pprn C r for 9 s limestones, whcrcas Turcki:in and Carr (in press) find an avcragc of 1 I ppm for three carbonate rocks an:ilyzctl b y iicuiroii activation and Hirst and l\iicliolls (135s) rcport an ;ivcragc of S pprn for ciglit British liincstoncs by a spectrographic tccliniquc. \Vc tisc the avcr;igc of these thrce scts o f clarn. I I ppin Cr. Tlic sandsronc valuc of 35 ppin Cr is tlic avcragc of Frolilich's 53 sand- stoiic and qiixtzitc samples. Turckian and Carr (in prcss) rcport an avcr:ixc of 7 ppm Cr for t\vo clctcrmiii;itions by iicutroii activation. Tlic pclqic-clay valuc is from Golclberg and :\rrlicnius (19%) :uid Frolilich (1959), and the carbonate dcep-sca-sediment value is from Turckian and Fccly (l35G).

h~ltrngonese: Tlic igncous-rock clatn are lrom Nockolcls (19%) and Green ;ind Poldcrvaart (1955). Thc shale value is the average of the values reported by Shaw (I%+) , Tourtelot (1957), and \Vcdepolil (1960). Ostrom (1957) reports an average of 1400 ppm A l i i for Car- bonifcrous limestone s3mpIes from Illinois, and Runncls and Schlciclier (1956) report an average of 550 ppm h l n for some Kansas limestones. \Ve use thesc two values to arrive a t an average of 1100 ppm hln, which is higher than most previous estimatcs. The sandstone order-of- magnitude value is a guess. T h e pelagic-clay value is from Goldberg and Arrhenius (1958) and IVcdepohl (1960). and the carbonate deep- sea-sediment value is from P. J. Wangersky (1958, P1i.D. thesis, Yale Univ.). Both,Correns (1937) and Wedepohl (1955) rcport that the clay-frce deep-sea carbonate tests contain about 200 pprn h.111.

T h e igncoiis-rock datn ;irc froin Neck- Iron:

. . . . .

' . . . . CHOICE

olds (1954) and Green and Poldervaart (1955); the sedimentary-rock data are from Clarke (1924); the pelagic-clay value is from Goldberg and Arrhenius (1958); and the carbonate deep- sea-sediment value is from P. J. Wangersky (1958, Ph.D. thesis. Yale Univ.).

Cobalt: Most of the values are from Carr and Turekian (in press), who used a combined neutron-activation and spectrographic technique to analyze a large number of specimens. The ultrabasic value is similar to the value obtained from the weighted average of ultrabasic. mineral analyses by Ross, Foster, and Myers (1954) in the manner described under chromium. T h e basaltic-rock value agrees with the average of 72 basaltic rocks by Nockolds and Allen (1956), Ahrens' (1951) value for North American rocks, and Smalcs, Mapper, and \Vood's (1957) value for oceanic islands. From Sandell and Goldich's (1943) data, 13 low- calcium granitic rocks give a n average ot 2.i ppm, whereas two high-calcium granitic rocks give :in avcragc of 5.Y ppm. Tlic syenite value is a guess based on Gordon and h.Iurata's (1952) d;tta. Tlic shale value is similar io that o i S11:tw (1954). Hirst and Nicliolls (19%) hiid an avcragc of 10 ppm lor cight British limestones, which ive coiisitler too high. The pelagic-clay v a h c is ;in avcragc of high valucs for die ['acific and low values for the :\tlantic (Golilbcrg and :\rrlicnius, 195s; Smalcs, Mappcr, and \\wcl, 1957; I-Iutcliinson rt af., 1955: Wcilcpohl. 1960). I f the carbonate deep-sea-scdiment value is based on a 90 per cent C:tCOX Atlantic Equa- torial core, a calculation from the data of

tppcr, and \Vood (1957) on a 67 per ccnt CaC03 Atlantic Equ:ttorial core containing an ;iverage of 11 pprn Co will givc 4 ppm. which compares with the value of 6 ppm obtained by Carr and Turekian (in press) for dcep-sca carbonate sediments.

The ultrabasic value was gotten in the iiiaiiiier described undcr chromium from tlic data of Ross, Foster, and hIyers (1954). The basaltic value is from Turekiiin (1956). The granitic-rock values are derived from Sandell and Goldich's (1943) data on 13 low-calcium granitic rocks and two high-calcium rocks. The sycnite value is a guess based on the few scattered data on this rock type (Sahama. 19-15: Butler, 1954; Gordon and Murata, 1952). The shale value is thc intermediate between Shuw's (1954) :tverage of 64 ppm and Turckian and Carr's (1960) average of 71 ppni. The sandstone \raluc is from Saliania's (1945) data on Finnish qurtzi tcs . Hirst a n d Nicholls (1955) report a n

h!ic/<e!:

OF DATA 181

average of 27 pprn N i in eight limestones. Runnels and Schleicher (1956) report 10 ppm for some limestone from Kansas, and Wede- pohl reports (unpublished) 25 ppm on a limestone composite made by Goldschmidt. A value of 20 ppm is chosen for'limestones con- sidering these data. T h e nickel values in pelagic-clay and carbonate deep-sea sediments are based on the same type of argument used for cobalt, and the values derived are from the same sources.

Copper: T h e ultrabasic value is based on a dunite from St. Paul's Rock in the Atlantic Ocean analyzed by Smales, Mapper, and Wood (1957) by neutron activation. Two peridotites of Moritg (1955) have Cu up to 20 ppm. The basalt value is from Turekian (1956) and corresponds with that of hlorita for Japanese rocks. The granitic values are derived from scveral scattered sources and represent the best estimate possible lrom tlie mmner i n which the data are reported. The soiirccs are: S:u~dcll and Coldich (19-t3), North Amcrican granitic rocks; S u p v a r a and Morita (1950) and I<uroda (l957), Japanese granitic rocks; and S m a h (195 j), analysis of C-1 granite. T h e syenite valuc is a guess.

Tlie sli;tlc value is the avcr3ge of sis sets of d:m: Sh:tw (1954). Littlcton formation. Dc- vonixi, IS ppm: Dcgcns. \ViIliams. and Keith (1957), Car1)onifcrous slialcs of Pennsylvania, 73 ppni; Turckian (unpublished), Fos I-Iills formation, Cretaceous. IS ppin: Sugawara ancl Morita (1950), Mesozoic of Japan, 55 ppm. Paleozoic of Japan, 40 ppm, and Paleozoic of Europc, 65 p p m - a l l composites. I-leide and Singcr (1950) report 105 pprn for the Rijt shalc of Jcna. N o data are available for sandstones. T h e x indicates the probable order of magni- tudc for sandstones. The limestone value is from three composites (93 srrmples) of German limestones of Paleozoic and Jlesozoic age (Wedepohl, 1955) averaging 2 ppm Cu and from data on the calcareous portions of'rocks from the Hanover mining district, New blesico, reported by Barnes (1957) with 5 pprn Cu. Heide and Singer (1950) report 5 ppm for the Muschelkalk limestone of Jena.

The deep-sea-clay and -carbonate data are from the same sources as cobalt and nickel, and the same method of calculation is used.

Zinc: T h e igneous-rock data are from Wede- pohl (1953 and unpublished). Sandell and Goldich (1943), hforita (1955). and Tauson and Pevtsova (1955), and agree in general with each othcr quite well. Tlie syenite valuc is from

182 TUREKIAN-.AND WEDEPOHL-DISTRIBUTION OF ELEMENTS

Morita. T h e shale value is from Wedepohl (1960). Sugawara and Morita (1950) report an average of 110 ppm on the same composites used in the copper determination; Heide and Singer (1950) report a value of 103 pprn for the R o t shale of Jena, and Barnes (1957) reports 70 pprn for limeless shales near Hanover, New Mexico. The sandstone value is from Wedepohl (1953), and the limestone value is the average of Wedepohl’s and Barnes’ data. T h e pelagic- clay value is from Wedepohl (1960), and that for the deep-sea carbonate is from unpublished data by the Same author.

Gallium: Sandell (1949) reports 1 ppni Ga for ultrabasic rocks, whereas Borisenok and Saukov (1960) report an average of 2 ppm. Wc have used the average of these two values. The data of the different workers on the Ga concentration of the other igneous rocks agree on the whole, but subtle differences exist. For basaltic rocks Borisenok and Saukov report 15 ppm Ga, and C. IC. Bell (1953, P1i.D. tliesis, hiass. Inst. of Tcchnology) reports 17 ppm Ga, whcreas Flcischcr’s (1955) average of a Iargc number of data from different workcrs on rocks ranging in 45-55 per cent Si02 is 20 ppm Ga. We choosc the average valuc of 17 pprn Ga for bnsalts. For high-calcium granitic rocks Bori- senok and S:iukov report 16 ppm, whcreas Bell rcports 17 pprn and Fleisclicr reports 20 ppm as an average of rocks ranging from 55-65 pcr cent SiO?. We use a value of 17 which is similar to the basaltic value. The Io\v-caIcium granitic rock values range from 16.5 pprn (Flcischcr’s average for rocks with greater than 65 per ccnt SiOz) to Bell’s 17 ppm, to Boriscnok and Saukov’s 19 ppm. We choose 17 ppm Ga for this rock type. The syenitic rocks range from 20 ppm (Bell) to 40 ppm (Borisenok and Saukov), and we choose a value of 30 ppm.

The shale value is from Bell, Sham (1954), and Wedepohl (1960), who find the same average value. Bell reports an average of 1.1 ppm Ga for four sandstones and 6 pprn for a quartzite, giving an average of 12 ppm. T h e limestone value and the carbonate deep-sea core valueare fromestimates in Goldschmidt (1954). Borisenok and Saukov report 10-30 ppm Ga for “marine oozes.” The pelagic-clay value is from Goldberg and Arrhenius (1958) and Wedepohl (1960).

Germaniirm: The data are from El Wardani (1957) and Onishi (1956), who generally agrce. El Wardani gives about 1.2 ppm for shalcs, whereas Onishi (1956) gives 2 ppm. We have chosen an intermediate value. Thc limc-

stone and carbonate deep-sea-sediment values are calculated on the basis of 10 per cent clay fraction. Pure Globigmna tests have 0.0 pprn G e (El Wardani, 1958). T h e pelagic-clay average is that of El Wardani; Onishi reported 1.6 ppm Ge.

Arsenic: The data for arsenic are all from Onishi and Sandell (1955a). They report that silicic volcanic rocks and serpentines have about 4 ppm As, which is considerably higher than the value for the usual igneous-rock types. Correns (1937) reports 7 ppm As for six samples of Atlantic pelagic clay.

Selenium: T h e values of all units escept the sediments are calculated from the sulfur abundances using a S to Se ratio of 6000, reported by Goldschmidt and Strock (1935) and Goldschmidt (1954), and have order of magnitude of significance only. T h e shale value is from Minami (19353). Sandstone and limestone values are those of Goldschmidt and Strock (1935). The deep-sea-sediment values are from Edgington and B y e s (1942).

Bromine: Tlic data are all from Bcline (1953). i\ssuming that all the bromine of the dcep-sca scdimcnts is in the S C ~ salt (Globigerinn shell samplc 6 I ppni Br) for the dcep-sen carbonate, we Iiavc ;itso used Bchnc’s pclagic clay value.

Rtrhidiirm: The ul t u basic v a h c is c a h - l a d to givc a I<b/K ntio thc samc as in basalts. AI ~ h c othcr igncous-rock values and thc shale and sandstone values are from Horstman (1957). Isotope-dilution analyses of composite basalts and granitic rocks by Cast (1960) give very close to thc same values. The carbonate deep- sea-sediment value is from Smales and Salmon (1955) and Horstman (1957), and the pelagic- clay value is from Wedepohl (19GO) and Horst- man (1957) and an estrapolation of Smales and Salmon’s (1955) data on deep-sea-carbonate- sediment samples with varying amounts of clay.

The ultrabasic value is bascd on the data of Pinson, :\hrens. and Franck (1953): the pclagic-clay value is from Goldberg and :\rrhenius (1958) and Wedepohl (1960) for carbonate-free sediment. .\I1 other values are from Turekian and Kulp (1956).

The data for yttriumwill influence the values for the rare-earth elements, since little or no information is available for most of the rare-earth elements in the common rock types.

The ultrabasic yttrium value is not known but is probably of the order of tenths o f a part per million. The basaltic value is from 72

Srronhm:

Yttrium:

.. . . . . . . . . CHOICE OF DATA. ., - . . .:-.. 183. . . . . . .

. . . . . - . .

basaltic rocks reported by Nockolds and Allen (1956). A confirmatory unpublished value for a one- to-one average of olivine basal ts and tholeiitic basalts is 25 pprn Y (Wedepohl). The granitic values are from Fleischer’s (1955) compilation, omitting the data of Laboratory #4 of his tables. The values are the average of the maximum and the minimum values he reports. Wedepohl (unpublished) gets 43 ppm Y for 17 German granites and (3-1. The syenite value is an average of Wedepohl’s 22 nepheline syenites and Butler’s (1954) value for a Norwegian quartz-free syenite having 20 ppm Y . Gordon and LMurata (1952) report a value of 130 pprn for a n Arkansas nepheline syenite. The shale value is from Minami’s (1935b) data. Wedepohl (1960) reports about the same figure. The sandstone value is a guess. Sahama (1945) reports a low value of 2 ppm for quartzite, but liis figures for granites (< 10 ppm Y) and syenites (< 10 pprn Y) are also low. The reason such a high value was chosen is bccausc most resistate dcposits must contain n consiilcrablc amount of rcsistant rare-earth mincrals. Onc can comparc yttrium in this sense with zirconium, for wliicli data are availablc. I f wc assume that the ratio oEyttrium i n sandstones rclative to shalcs is thc same as for zirconium. we get 40 ppni Y. Wcdcpohl’s un- piililishctl rcsirlt for rhrcc limestone composites (93 snmplcs) is 30 ppm Y. Tlic pelagic-clay valuc is irom Golilberg and Arrhcnius (1955) and \Vedcpohl (1960). T h c carbonate deep- sea-scdiincnt value is that of Wedepohl’s un- publislicd data on the ;\tlantic.

Tlie da ta for all rock types cscept the dccp-sca clays are from Degcn1i:irdt (1957). His values generally agree with thosc of other workers. Tlie pelagic-clay value is from Goldberg and Arrhenius (1955) and IVedcpolil (1960).

Very littlc information on the hafnium abundance in common rocks is available. We can, ho\vcver. use the coherence of hafnium and zirconium as a method of estimating the hafnium content of rocks from the zirconium content. Tlie Zrj’Hf ratio varies in zirconium with the rock type. If zircon and the mafic minerals are the main carriers of both the zirconium and hafnium, and if the Zr/Hf ratio of coexisting zircon and the mafic minerals is the same, then a method of approximation is avail:rblr to us. Gottfried, Waring, and \Vorth- ing (1956) and Kosterin, Zuev, and Sheva- leevskii (195s) have determined thc Zr/Hf ratio of zircon from various rock types. The

Zirconiirrn nnd h;!/niirm:

~~~

only difference in the two sets of data is in syenitic rocks where Gottfried, Waring, and Worthing report a higher ratio. Table 1 gives an estimate of the hahium content of the . various rock types based on the Degenhardt and Kosterin, Zuev, and Shevaleevskii data.

.

TABLE EST ESTIMATION OF HAFNIUM COSCENTRATIOSS IN “IGNEOUS” ROCKS

Zr /Hf (in zircons)

pprn Zr (Kosterin, (in rock) Zuev, and pprn Hf

(Deynhardt) Shevaleevskii) (in rock)

Ultrabasic rocks 45 70 0.6 Gabbro I40 70 2 .o Granodiorite 1.10 60 2 . 3 Granite I i5 45 3.9 ’ Syenite 500 45 ’ 11.1”

(Sedimentary rocks Zr/Hf=5S)

Cooley et a f . (1953) providc data which give 2 Zr/Hf averagc ratio of 44, indicating perhaps that most of their zirconium numbcrs wcrc of granitic or sycnitic afinity.

A’io6iwn and tmruliim : The igneous- rock data of Ranknnia (1944; 194s) gcncrally ngrcc with thosc of Znamcnski (1957) csccpt for tlic high-calcium grmitic rocks. For tlicsc Itanknma reports 3.6 ppni Sb and 0.; ppni Ta Ibr six Scandinavian dioritic rocks. Znnnicnski Sets an average of 20 pprn N b and 3.6 ppm Tri for this rock type, and w e shall use liis values. Grimaldi (1960) reports 12 ppm N b for standard granitc G-I and 9.6 ppm N b for standard diabasc 1V-I . The scdirnentary-rock and deep-sea-sediment data arc from Rankama.

Mo~)6denirrn: The igneous-rock values arc thc averages of the data on each rock typc reported by Kuroda and Sandell (1954) and Vinogmdov, Vainshrein. and Pavlenko (195s). The latter’s values, Jctcrmined spectrograplii- cally on Russian rocks, are generally higher (except for ultrabasic rocks) than Kuroda and Sandell’s, determined colorimetrically. Ishi- mori (1951) reports a value of 0.9 pprn hlo for the average of 10 Japanese basalts, which is comparable to the low value Kuroda and Sandell derived for basaltic rocks.

The sedimentary-rock data are from K u r d a and Sandell (19%). They also report 3 ppni Mo for both carbonate and clay deep-sca sediments. Goldberg and .-hhenius (195S), lion.- ever, report 45 pprn for East Pacific pclagic- clay samples, and \\:edcpohl (1960) reports 9

184 TUREKI \N AND WEDEPOML-DISTRIBUTION OF ELEMENTS

pprn for Atlantic pelagic-clay samples. We use the average of these two last values, although we cannot explain the discrepancy with Kuroda and Sandell. We use Kuroda and Sandell’s carbonate deep-sea sediment value, however.

The fen new data available are from Vincent and Smales (1956), who determined palladium by neutron activation. T h e granitic-rock values are probably of the order of magnitude indicated.

All the silver data, where numbers arc listed, are from Hamaguchi and Kuroda (1959), who analyzed primarily Japanese rocks. Tlie unpublished data of .A. Kvalheim quoted by Goldschmidt (1951) appear to be too low. Tlie high values for diabases from Ontario reported by F:iirbairn. .Ahrens, and Gorfinkle (1953) are probably characteristic of that region only and not :ipplicablc generally.

basic rocks. b u c chc order of inagaitudc listcd is prob:ibly corrcct. Tlic igneous-rock data arc from Sandcll and Goldich (194.3). I’reuss (1940) got 0.2 ppm Cd for a graiiirc composite and 0.3 ppin Cd for :i shale coniposirc. Tlic clccp- sex-sctlimcnt data :irc cnlcu1;icccl Crom hlullin : ind liilcy (1956). Tlieir :ivcr:igc of rcccnt calcium carbon;itc shclls is 0.035 ppm which \vc L I X for the liincstoiic v;iluc :iltliough this is uiirlouhtcrlly a lowcr limit sincc other contributing phascs in :I norninl liincstonc have not been consiclcrcd.

Tlic data :ire I’rom Shaw (1952b).

Pnihdiirm:

. Siicw:

C d t ~ z i r o T z : N o data COCIICI be found for ultra-

I t d i r r t v : I-Io\vcver, rhcre are sonic unccrtnintics licrc. The high-calcium granitic rocks havc a value much loivcr than cithcr the lon-dciiim granitic rocks or tlic bmlt ic rocks. Wager, Smit, ; i n d Irving (195s) report ncutron-~icti\:ation vnlues for indium in \V-I stnndard diabase and the chill zone from the Skncrgaard comples. Greenland. ;is 0.064 and 0.0% ppm rcspective- ly . These numbers are lower than the values chosen for the table.

The tin data :ire primarily from Onishi and Sandell (1957). The higli-calcium granitic rocks \Yere arbitrarily assigned a value of 1.5 ppm. Degens. l\’ilIianis. and Keith (1957) give 3.2 ppm for the value of their Carboniferous shales, and \Vedepolil (unpublished) finds a valuc of 5 ppm for shales. T h e value of Onishi and Sandell for shales (1 1 pprn) has been aver- aged with the above authors’ to give the value in the tablc.

A n ~ i t n o n ~ : The data are a11 from Onishi and S;indcll (195%). These authors were not com- plctcly saiishcd with their results and claim

Tin:

they should be taken tentatively. In the ab- sence of any later information these are the best estimates available.

The estimates for all the rock types except the deep-sea-sediment data are based on the monograph “Geochemistry of Iodine” (Chilean Iodine Educational Bureau, 1956). The deep-sea-sediment data are based on the chlorine content of these sediments as inferred above and the I/C1 ratio of the sea (= 2.27 x lo-”). These values may be too low, since there is a correlation of organic content of shales and iodine, so some of the iodine may be enriched in sediments relative to the sea.

Cejilrtn: Thecesiumconcentrations in ultrabasic rocks, limestones, and sandstones are not known except that they are all probably less than 1 pprn (I-Iorstman, 1957). Tlie granitic value is calculated from Cast (1960), who analyzed two granitic composites. From the valucs he obtained for Li and Sr, it appears that tlic composites could be rcsolved into one part low-calcium to oiic part high-calcium granite rock with tlic assumcd valucs listcd to give the obscrvcd v31uc of 3.2 p p n . Horstinan (1957) rcports 1 ppin Cs as an :ivcr;igc of a composite of 6G samples. Tlic sycnitc vnluc is from the avcragc of tlircc sycnitic rocks from East Grcen1:intl :~nalyzcd 1)y Liebciibcrg (1956). Tlic slialc v:iluc is from I-lorstman. Canncy‘s (195Z)e avcragc sccms to be rather high. Tlie dcep-sea’ data arc froin Sinales and Salinon (1955). They report a value of 0.4 ppm for 90 per cent calcium carbonatc seclinient increasing to 1.5 ppm for thc portion of tlic corc analyzed which has SO per cent CiCOp. By extrapolation to 0 per ccnt calcium carbonate, :I value of about 6 ppm is obtaincd. This compares with Horst- man’s (1957) value for a pelagic-clay composite. The basaltic value is the a v e r q e of cesium values determined by Cast (1960) on three basalt composites by isotope dilution. Cabell and Sinales’ (1957) value for \V-1 standard diabase (1 .OS ppm Cs) agrees, whereas their value for the S k a e r p r d chilled marginal gabbro (0.10 pprn Cs) is low.

The value used for ultrabasic rocks is a single activation analysis of a dunite made by Hamaguchi, Rced. and Turkevich (1957); it is considcrably loxver than the 6 pp’n reported by Pinson, Ahrens, and Franck (1953). Von Engclhardt’s (1936) figures for olivine scatter around 1 ppni. Cast (1960) reports an average of 333 pprn for three basalt composites analyzed by isotope dilution. Hamaguchi. Rced, and Turkcvich (1957) get 310 ppm for a

Zodine:

Barikm:

IS5 - . . .

. . . . .-. CHOICE OF DATA

- . .

single activation analysis of a Hawaiian basalt. T h e basaltic value from Nockolds and Allen (1956) for 72 basaltic rocks is 180 ppm, considerably lower. The granitic values are calculated from Gast’s (1960) average granitic value of 620 pprn as described under cesium. The syenite value is the average of the values of von Engelhardt (1936) and of Sahama (1945). The shale value is the average of data from five difierent studies: Degens, Williams, and Keith (1957), Carboniferous shales, 450 ppm Ba; Tourtelot (1957), Pierre (Cretaceous) shale, 720 ppm Ba; Macpherson (195S), Precambrian graywackes, argillites. and low-grade schists, 440 ppm Ba; Shaw (1957), Littleton formation (Devonian) pelitic rocks, 5SO ppm; and Wede- pohl (1960) Japanese and European shales, 700 ppm. Tlic limestone valuc is based on modern molluscan shells (Turckian and Armstrong, 1960). The saiidsronc valuc is a giicss. Quartz gcncr:illy Ii:is very lu\v barium, hut the prcscnce of Iica\.y mincr:ils and BaSO4 cement will raise the Iigurc. iVc Iiave nssumcd that von Engel- 1i:irtlt’s \:;ilucs ol I70 ppm Ba for s;inclstoncs ;ind I20 ppin Lki lor limcstoncs ;ire too high. Tlic v;iluc Cor dccp-sca chys from the :\tlantic is 700 ppni (\\’cdcpohl. 1960) ;ind I‘rom the 1’:iciIic 4000 ppm ((;oldbcrg ; i i id Arrlicnius. 1358). Tlic :ivcr:igc o i thcsc valucs is used. Thc I’ncific 1i:is :I strong b:irium sulf;itc component cithcr dispcrscd or in the Corm of concrctions (Goldbcrg ;mil :\rrlicnius, 19%). Tlic carbon- arc deep-sca-sediment \.aluc is :in unpublished rcsult for four “G/obiyerinli ooze’’ cores from the Atlantic (Wedcphl) .

The iiltr;ihasic value is a guess. The basalt \ d u e is thcit for Ontario diabase from Fnirbairn. .-\lircns. ; i d Gorfinkle (1953). For granitcs, Kockolds and :\lien (1953), Sahama (l945), and .-\hrcns (1954) report :illout thc same avenges. arounci 55 ppm La. The graiiodiori tc averagc is from Nockolds and :\llcn (1953). The Lime authors report (1954) 95 ppm La in trachytes. whereas Sahama (1945) got 50 ppm in syenites. An intermediate figuye is used. Nepheline syenites are again much higher (Gordon and l lurata , 1952).

Wcdcpohl checked klinami’s (1935b) La determinations and found his shale averages too low (this concerns only the La values of his rcport. however). \Vedepolil (1960) reports 92 ppin La as an average of shales. which we have used. The deep-sea-clay valuc is again a one-to- one average of that for Pacific (Goldberg and Arrhenius. 19%): 130 ppm. and Atlantic samples (\Vcdcpolil. 1960) : 3s ppm. For the

Lan//itrnrrrn:

~

carbonate deep-sea-sediment, 10 per cent of the -clay value is assumed.

Other rare-earth elements: The other rare- earth values for most of the rock types are based on the assumption that the ratio of each of the rare-earth elements to yttrium is the same as it is in shales as determined by hlinami (1935b). Data on Ce, Pr, Nd, Sm, Eu, G d , T b , Dy, Er, and Yb in granites are reported by Sahama (1945). Except for a low cerium value of Sahama, there is a good agreement with our computed values. In pelagic clays from the Pacific, 100 pprn Nd and 12 ppm Y b could be estimated (\Vedepohl, unpublished), also in close correspondence to our prediction. A few small deviations from our valucs result if dne uses the considerations of llssuda (1957) as a basc for thc computations.

Tlic ultrabasic value is from Vino- gradov, Vainshtcin. and Pavlenko (1955). The ’

basaltic, granitic, :ind syenitic values :irc iiitcr- mcdi:ite bctwcen the :ivcragcs of thcsc autliors and those of S:incfell (I94G). In al l but the sycnitic rocks Saiiilcll’s values arc lo\vcr. Iclliry (1959) reports I O ppm for rhc avcragc of alk;ili rocks from Uganda. \Ve consider this too high for ;I general average for syenitic rocks. His avcragc lor grani tcs, however. (uii~~istiiisiiislicd as to low or high calcium). 1.4 ppm. is comparable to rhc valuc we have chosen. The sli:ilc and sandstone values arc the analyscs in;itIc by Vinogr:irlov, \’aiiislitcin. : i d Pavlenko of com- positcs. of scvcr:il thousand wnplcs of tliesc rock types prepared by Ronov. The limestone valuc is thc average of seven analyses on Alrican limestones by Jctlery. Cnlortunatcly no data are :is yet available for deep-sea sediments, and so order of magnitude gucsscs have been made.

Go/d: T h c only recent data :ire neutron- activation detcrminntions by Vincent and Crockct (1960) and Crocket. Vinccnt. and Wager (195s) on some ultrabasic rocks. basaltic rocks. and standard granite G-I. The values for these igneous-rock types are from their data. I t is assumed that the gold content of :ill other rocks will be of the same order of magnitude, although Clarke (1924) reported 0.03 pprn .Au for sandstones and 0.005-0.009 pprn for limestones.

Mercrrqr: The v:ilues for the igneous rocks where any are listed and shales are averages of the determination of Stock and Cucuel (1934) and Preuss (1940) on the same composites of German rocks prepared by Goldschmidt. The two sets of valucs agree Cidy w l l . Tlic limc-

Tirng.cren:

186 TUREKIAN AND WEDEPOHL-DISTRIBUTION OF ELEMENTS . .

stone value is the average of the determinations on the Muschelkalk by Stock and Cucuel (1934) (one analysis, 0.033 ppm Hg) and Heide and Bohm (1957) (average of several specimens, 0.048 ppm Hg). The latter authors also report 0.19 ppm for the underlying red shale (“Rot”) near Jena. The sandstone value is from a single analysis by Stock and Cucuel. The other values are order of magnitude guesses.

The data are from Shaw (1952a), Ishimori and Takashima (1955), and Preuss (1940). The reported value for ultrabasic rocks and syenites is from Shaw, that for basalts is the average of Shaw’s (0.13 ppm), Ishimori’s (0.3 ppm), and Preuss’ (0.3 ppm) data. T h e granodiorite value is intermediate between Shaw’s (0.43 ppm) and .Ishirnori’s (1.0 pprn). For a composite of German granites, Preuss got 3 ppm. Shaw reports 3.1 ppin as an average for granites and Ishimori 0.9 ppin for Japanese granites, resulting in an average of 2.3 ppm TI. Preuss and Ishimori get the sainc v:iluc for the ame composite of European c;1rbo11accot1s shales, 2 ppm TI; Sliaw reports a shalc average of about 0.S ppm. \Vc use thc average of these two. Canney (1952) reports a value of about 0.4 ppm TI for shales. For Pacific pclagic cl;y Shaw reports 1.2 ppm TI, whereas Atlnntic clay has 0.42 ppm TI. Rced, Kigoshi, and Tiirkevich (1955) report a value of 0.94 ppin TI for a perthite from a graniic and of 0.07 ppm TI for a basalt using I I C ‘ L I ~ ~ O I I activation. Tliese numbers are slightly lower than Sliaw’s.

Lcnd: The ultrabasic value is based on the nnge (-0.1 to 0.01 ppm) reported by Tilton and Rced (1960). Tlie rest of the data are primarily from Wedepolil (19%). The ;ltlantic pelagic clays have a value of 45 ppm (Wedcpohl, 1960), the Pacific clays :I value of 110 ppm. Goldbcrg and Arrhenius (19%) got 140 ppm for the Pacific. Hence again there are difier- ences in the chemistry of the sediments of the two oceans. The average of the two is used. The carbonate deep-sea-sediment value is 10 per cent of the average pelagic-clay value as be- fore. Turekian and Feely (1956) report a mean value of 4 ppm for an Atlantic Equatorial arbonate core. The granitic-rock data are con- firmed by ;\litens (1954) for Nor th American rocks of lon.-calcium :ind high-calcium granitic composition. Gordon and Murata (1952) report a value of i ppm for an Arkansas nepheline syenite. Shaw.(1954) lists i v a l u e of 16 ppm for pelitic rocks. \vhich compares with Wedcpohl’s data. Degens, \\’illiams, and Keith (1957) on the other hand tind 35 p p n for Carboniferous

Thallium:

shales of Pennsylvania. Heide and Lerz (1955) report a value of 2 I pprn for the Rot shale near Jena and 7.9 ppm for the Muschelkalk limestone.

Data for granites and shales are from Preuss (1940). Reed, Kigoshi, and Turke- vich (1958), as a byproduct of their work on meteorites, have published Bi values for a perthite from a granite and a basalt (from the Snake River region, U.S.). These numbers are included in Table 2 on the premise that some idea of the possible value for the abundance of Bi in a rock type may be better than no idea. There are other values in the literature, but they all appear high. Preuss (1940), for example, reports 2 pprn Bi for a composite of German granites and 1 ppm Bi for shales. Brooks, Ahrens, and Taylor (1960) report a wide range of values for a variety of rocks in a preliminary report.

T h e ultrabasic value for thorium is ticrived from the uranium value (~vhich is used for this rock type) of a single dunite by Hamaguchi, Reed, and Turkevich (1957) and assuming a T h / U ratio of 4. These low values h a w also been found for chondrites. Tlic b:is;iltic and syenitic valucs are from Evans and Goodman (1941). T h e granite values are from Whitfield, Rogcrs, and ildams (1959). Tlic shalc and limestone valucs are from Adams nnt l \Vcaver (195s) ;inti tlic sandstone value from hlurrcly ;inti .-\dams (1955). The uranium value for dcep-sea-clay sediments is derived from Starik ct al. (195s). On a clay core from the southern part of the Indian Ocean they find an average of 1.3 ppm U.

For a core with about 40 per cent CaCol these workers found the same value for uranium and a high value for thorium (13.5 pprn). Others have found very low concentrations of uranium in the carbonate fraction (<<I ppm U. \V. S. Broecker. personal com- munication) indicating that the uranium value for average carbonate deep-sea sediments is the order of magnitude indicated in Table 2 rather than the high value mentioned above. Picciotto and IVilgnin (1954) report a n average of about 5 pprn T h for a Ccntral Pacific pelagic-clay core. The average between this and the higher value obtained by Starik and his coworkers on Indian Ocean sediments serves as ou r choice for the abundance of thorium in pelagic clays. The carbonate deep-sea-sediment values for T h from the literature are widely divergent, and so we indicate only an ordcr of magnitude estimate.

Bismuth:

Thorium and uranium:

- -~~

. . . . . . . : .

I

\ I

CJ 25,OO 1. " . . I : P .

REFERENCES CITED 187

. . REFERENCES CITED Adanis, J. A. S., and Weaver, C. E., 1958, Thorium-to-uranium ratios as indicators of sedimentary proc- .

esses: .Example of concept of geochemical facies: Am. Assoc. Petroleum Geologists Bull., v. 42,

Ahrens, L. H., 1951, The lognormal distribution of the elements: Geochim. Cosmochim. Acta, v. 5,

Barnes, H. L., 1957, Trace-element distribution in shales near Hanover, New Mexico, mining area (Ab-

Behne, W., 1953, Untersuchungen zur Geochemie des Chlor und Brom: Geochim. Cosmochim. Acta

Boriskenok, L. A., and Saukov, A. .4., 1960, Geochemical cycle of gallium: XXI Internat. Geol. Congress, Copenhagen, Part I, p. 96-105

Borodin, L. S. , 1956, On the distribution of beryllium in the Khibina alkalic massif and on the clarke of beryllium in nepheline syenites: Doklady Akad. Nauk SSSR, v. 109, p. 811

Broecker, W. S . , Turekian, K. K., and Heezen, B. C., 1958, The relation of deep sea sedimentation rates to variations in climate: Am. Jour. Sci., v. 256, p. 503-517

Brooks, R. R., Ahrcns, L. H., and Taylor, S. R., 1960, The determination of trace elements in silicate rocks by a combined spectrochemical-anion exchange technique: Geochim. Cosmochim. Acta, v. 18,

Butlcr. J. R., 1954, The geochemistry and mineralogy of rock \\feathering. (1) The Nordmarka area,

Cabell, M. J., and Sinalcs. A. :I., 1957, Thc determination of rubidium and caesium in rocks, minerals

Canncy, F. C., 1952. Some aspects of the geochemistry of potassium, rubidium, cesium, and thallium in

Carr, \I. 13.. and Turekian. K. ti., in press, ’fhc geochcmistry of cobalt: Gcochim. Cosmocliim. :\cta Chilexi Iudinc Edt~cational R u r c w 1933, Gcochemistry of lodinc: London. Stonc House, 150 p. Cl:irkc, I:. W., 1924, Data of gcochcmistry: U. S. Geol. Survcy Bull. 770, S i 1 p. Coolcy, R. A., Mirtin. A. V., Fcldinan. C., 2nd Gillcspic. J., 1953. The I-If to Zr :ibuntlaiice ratio and

spccific radioactivity of some ores: Geochern. Cosniochirn. :\cra, v. 3, 11. 30-33 Corrcns, C. W., 1937. Die Scdiincntc des iquatorialcn :\tlnntischcn Ozeans. 7 : \Viss. Ergcbnisse d.

Ileutschcn :\tlant. Expedition Atctcor 1925-27, v. l11/3, Berlin - 1956, Tlic geochemistry of the halogens, p. 183-23-1 in Ahrcns, L. H., Rankama, K,, and Runcorn,

S. ti.. Physics and chemistry o l thc Earth. volumc I : London, Pcrgmnon Prcss. 317 p. Crockcr. 1. H., Vinccnt. E. ;\., and \V;iger, L. R.. 195S, Thc distribution of gold in some basic and ultra-

basic igneous rocks and minerals: Gcochiin. Cosmochim. ;\ctn, v. I t , p. 153-151 Degenliardt, 13.. 195;. Untersuchungen zur geochemischcn Vcrteilung dcs Zirconium in der Lithosphire:

Geochim. Cosmochim. Acta, v. 11, p. 279-309 Degens. E. T., Willinms. E. G., and Keith, X.1. L., 1957, Environinental studies of Carboniferous sedi-

ments. part I : Geochemical criteria for differentiating marine and fresh water shnlcs: Am. . h o c . Petroleum Geologists Bull., v. 11, p. 2127-2155

Edgington. G.. and Byers. H. G., 1912. Geology and biology of North :\!Inntic deep-sea cores between Nen.foundland 2nd Ireland. Part 9: Selenium contcnt 2nd chemical analyses: U. S. Geol. Survey Prof. Paper 196-I.‘. p. 151-155

El \\‘xdani, S. :I.. 1997, On the geochemistry of germanium: Geochim. Cosmochim. ;\eta. v. 13, p. 5-19 - 1958, Marine geochemisrry and the origin of Pacific pelagic clay minerals: Geochim. Cosmochim.

Engclhardt. W. von, 1936. Dic Geochernie des Barium: Chcmie d. Erdc, v. 10, p. ISi-1-16

p. 387430 1 .

p. 49-73

stract): Geol. SOC. America Bull., v. 68, p, 1699

V. 3, p. 186-21i

p. 162-175

Oslo: Geochim. Cosmochim. i\cta, v. 6, p. 265-281

: i d inetcoritcs by ~ieutroii.ncrIvation analysis: Analyst, v. 82, p. 390-405

scdiinents (Abstract): Geol. Soc. :\mcrica Bull., v. 63, p. 1238

Acta, v. 15, p. 137-254

Evans. R. D., :ind Goodman, C., 1941, Radioactivity of rocks: Geol. SOC. .\merit? Bull., v . 5 2 , p. 199- 190

Fairbairn. 13. \V., ;\hrcns. I,. I-I., nnd Gorfinkle. G., 1953, Minor element content of Ontario dinbase:

Fleischer, X l . . 1953. Keccnt estimates of the nbundnnccs of the elements i n the Earth’s crust: U. S. Gcol. Geochim. Cosmochim. :\ct:i, \.. 3, p. 3 4 4 6

Survey Circ. 285. 7 11.

TABLE ~.-DISTRIBUTION OP TI18 ELELIENTS IN T l l E FIARTII’S CRUST (lixpresscd in parts per inillion) *

“Igneous” Rocks Ultrabasic Baultic Granitic Rocks Sycniws

Rocks I-Iigli I.ow Calcium C:ilcium

I I.lvilrogcn 2 lleliuin 3 Lithium 4 Ijeryllium 5 Horon 6 Carbon 7 Nitrogcn S Oxygen 9 Fluorine

10 Neon I I Socliuni I 2 hhgncsium 13 Aluminum. I 4 Silicon 15 Phosphorus 16 Sulfur I7 Clilorine I8 Argon I9 I’otnssiuin 20 Cdciiiin 21 Scindium 22 l’ iuniuni 23 Vm:iiliiuii ’

2 1 Clironiiuin 25 Al:uig:incsc 26 Iron 27 Cobdt 28 Nickel 29 Copper 30 Zinc 31 C;illiuni 31 Gcrmaniirm 33 Arsenic 3i Sclcnium 35 I3rominc 36 Krypton 37 I<iibid i uni 36 Strontium 39 Yttrium 40 Zirconium 4 1 Niobium 4 2 hfolyWcnum 43 Technctium 4 4 Ruthenium 4 5 I<licKlium 46 kill:idiuni 47 Silvcr 48 Cadmium 49 Indium 50 Tin 51 Antiniony 52 Tellurium 53 Icwlinc 54 Xcnon 55 Cesium 56 Rariutn 57 lanthanum 58 G r i u m 59 I’rascalymium

-iTNiX$iTm--- 61 l’rnmethih& ., 62 Sam3rium 63 Europium 64 Gxloliniurn 65 ’l‘d)iiun 66 Dysprosium 67 I4olniiuin 68 Erbium 69 Tlluliuin 70 Ytterbium 71 I.utctiuin 72 Hafnium 73 Tantalum 74 Tungsten 75 Rhmiiiin 76 Osniiiinl 77 Iridium 78 1’l:itinum 79 Gold 80 hfcrcury 81 Tlialliiun 82 Lesd 83 Ijismutli 84 Poloniiim 85 hstatinc 86 Radon 87 Francium 88 Radium 89 Actinium 90 Thorium 91 Protactinium 92 Uranium 93 Neptunium 94 Plutonium

I I I le 1.i Be I3 C N 0 F Nc N3 M 6 A I Si P S CI Ar K C a s c T i V Cr hln i:c co Ni CU Z n Ga Gc .4 s Se R I Kr Ilb Sr Y zr N b hto T c Ru I<ll I’d ;\ g Cd In So s I) l‘C

I xc c s Ua 1.3 C e

A B 0.x 0 . X 3.

A 6. A

100. I3

4200, 204,000. zo,ono.

205,000. 220, 300. 85.

I 3 40

15. 25,000.

300, 40.

1600. 1620.

94.300. 150.

2nnn.

50. IO.

I . 5 I . 5 I . 0.05 I

I%

1. 0.x

45. 16. 0 .3 C D 1) 0.12

0.x 0.01

0. I D

13 0 . X 0 .4

0.2

0.06

0 . 5

0.5

0.x 0.x

~ . _ _ _ L ‘ - . o x P r 0.x

I’m C Sm 0.x I‘u 0.x . Gd 0.x T b 0.x

0.x 0.x

DY 1-10 1:r 0.x T m 0 . S Yb 0.x Lu 0.x 11 f 0.6 T a I .o W 0.77

OS I) RC n Ir I ) I’t 1) Au 0.006 1 k 0.ox TI 0.06 Ph 1 . Bi D PO E At E Rn E F r E Ka E Ac I: T h 0.004 Pa E U 0.001

1: F Pu

NP

A B 17.

I . 5. A

20. A

400. I3

i8,ono.

78,000, 46,000.

230,000. 1100. 300. 60. B

8300 76.000.

30.

250. 170.

86,500. 48.

130. 87.

105. 17. 1.3 2 . 0.05 3.6

I3 30.

465. 2 1 .

14Q. 19.

I .5 C D D 0.02

0.22 0.22 I .5 0.2

D 0.5 13 1 . 1

330 .- 15. 4s.

13,800.

1500.

n.11

4.6

C 5.3

.8 5.3

.8 3 .8 1 . 1 2.1 0.2 2.1 0.6 2.0 1 . 1 0.7 n I’) I ) D 0.004

0.21 6. 0.007 E E E E I: E 4 . II I . F F

----20: -. -

0.09

A I3

24. 2 . 9 . A

20. A

520. B

28.400. 9inn.

82,onn. 314,000.

920. 300. 130.

I3 25,200.

1 4 .

88 . 22.

540.

7 . 15. 30. 60. 17.

25,300,

3400.

29,600.

I .3 I .9 0.05 4.5

II

440. 35.

140. 20.

I .o C D D 0.oox 0.051 0.13 0.ox I .5 0 .2

I)

I3 2 .

4 5 . 81.

7.7 33. - C 8.8 I . 4 8 .8 1 . 4 6 .3 I .8 3.5 0.3 3.5 1 . 1 2.3

’ 3.6 I .3

D I) I ) I)

0.08 0.72

15. D E E E li 1: I: 8.5 E 3.0 I: F

110.

n.5

420.

.- ..

n .oo4

A I3

40. 3.

I O . ,I

20. .A

850. I3

25.800. 16on.

72,000. 347,000.

600.

Zoo.

5100.

1200.

300.

I I 41,000.

7.

4 4 .

390, 4 . 1

14,200.

in.

I .o 4.5

39 . 17.

I .3 I .5 0.05 I . 3

I3 170. loo. 40.

175. 21.

C I ) I)

I .3

n.oox 0.037 0 . 1 3 0.26 3. 0 .2

I)

1% 1.

840. 5 5 . 92.

8 . 8 - 37.

C 10.

I O .

0 .5

I .6

I .6 7.2 2.0 4.0 0.3 4 . 0 I .2 3.9 4 . 2 2.2

I) D I ) I ) 0.004

2.5 19. 0.01 E IS E 1: E E 17. E 3 .O 1: 1:

0.08

A n

28. 1 . 9. ;\

A

I I

30.

1200.

40.400.

as,nno. 291 ,nnn.

800 .

520.

48,000. I 8,nno.

30.

5800.

300.

I3

3. 3500.

2. 850.

36,700. I . 4 . 5.

130. 30.

I . I .4 0.05 2.7 I3

110. 200. 20.

500. 35. 0.6 C D

I’) 0.ox 0 . I3 0.ox

X . 0.x

I)

n 0 .6

n

n.5

1600. 70.

161. 15. GF.-. - > C 18. 2.8

IS . 2.8

13. 3.5

0.6 7.0 2.1

2.1 I .3

1) I ) 1) I ) 0.oox 0.ox 1 . 4

7.0

1 1 .

12. 1) F. E E I: IS I: 13. E 3.0 1’ F

Sedimcntuy Rocks Shalcs Sandstoncs Carbonates

A B

66. 3.

100. A A A

740. R

9600.

80,non. 73,oon.

700. 2400.

180.

26,600. 22,100.

13. 4600.

130.

15,000.

13

90. 850.

47,200. 19. 68. 45. 95. 19.

I .6 13. 0.6 4 .

I3 140. 300. 26.

160. 11. 2.6 C D 1) I) 0.07 0.3 0.1 6.0 I .5

1) 2.2

I3 5.

580. 92. 59. 5.6

2K. ~ ’

C 6.4 I .o 6.4 I .o 4.6 1.2 2.5 0.2 2.6 0.7 2.8 0.8 1.8

D I) I) I ) 0.oox

I . 4

I ) E E E E I. I. 12. E 3.7 F F

0.4

20.

A

IS . 0.x

35. A A A

270. B

3300.

n

7000. 25,000.

368,000. 170. 240.

10. I3

10,700. 39,100,

I . 1500.

20. 35. xo .

9800. 0.3 2 . x . 16. 12. 0.8 1 . 0.05 I .

I3 60. 20. 40.

220. 0 .ox 0 . 2 C D D D 0.ox 0.ox 0.ox 0.x 0.ox

I’) I .7

B 0.x

xo . 30.

I 92.

A

5. 0.x

20. A A A

330. I3

n

400. 47,000.

4200. 24,000.

400.

150.

2700. 302,300.

400.

1200.

B

I .

20. 1 1 .

1100. 3800.

0 .1 20.

4 . 20.

4 . 0.2 I . 0.08 6 .2 I? 3 .

610. 30. 19. 0 .3 0.4 C D D D 0.ox 0.035 0.ox 0.x 0.2 D

I . 2 R 0.x

10. X . 11.5

1 . 1

i’C C 10. I .3 1.6 - 0.2

10. I .3 I .6 0.2 7.2 2 .o 4.0 0.3 4.0 1.2 3.9

I .6 o.nx

I) 1) 1) 1) 0.oox 0.03 0.82 7.

D E E

0.9 0.3 0.5 0.04 0.5 0.2 0 .3 0.ox 0.6 D 1) I) I ) 0 . oox 0.04 0 .ox 9.

1) 1: E

E E

0.45 2.2 r: 1’ F 1:

Deep-sea Sediments Clay ’ , Carbonatc

A

5. 0.x

55. A A A

540. B

20,000. 4000.

20,000. 32,000.

350. 1300.

n

21,onn.

3 12,4110.

B 2900.

2. 770.

20. 11.

1000: 9000.

7 . 30. 30. 35. 13. 0.2 I . 0.17

70. B

IO. 2000.

42. 20. 4.6 3. C D D D 0.ox 0.ox 0.ox 0.x 0.15 D 0.05 R 0.4

190. IO . 35. 3.3

f

C’ 3.8 0.6 3.8 0.6 2.7 0.8 1.5 0.1 I .5 0.5 0.41 0.ox 0.x u I) D D 0. oox 0.ox 0.16 9.

D E E E I: E E X. E 0.x F F

A B

57. 2.6

230. A A A

B 1300.

40.000. 21,000. 84,000

250,000. 1500 1300.

2 1,000. B

25,000 29,000.

19. 4600.

120. 90.

6700. 65,000.

74. 225. 250. 165. 20. 2.

13. 0.17

70. ’

B 110. 180. . 90.

150. 14. 27. C D D D 0.11 0.42 0.08 1.5 I .o

0.05 ,

R

D ,

6. 2300.

115. . 345. ,

33. .-. . { 40 -. . _- --.

38. . 6 . ”

38. 6.

27. 7.5

15. . 1.2 ,

15.

c’ . ,

4.5 4.1 ’ . 0,s

D 1) 1) :’

I) 0.oox ’

x . ’ , ’

0.x , . 0 .8 ,, ’

so. 11 E I .

E E E

E 7

E .

I’

* In somc w x s , only order of magnitude cstimatcs could LK made. T h c x arc indiwatcd by the symbol X.

A: These elements arc thc basic constituents of thc biosphere, hydrosplierc, and at- moyhcre. Oxygen is also the most important clement of tlic litlicnpherr, wliercar carbon is important in scdimcntary rock.

B: Tlic n r c gases occur in tlic atmosphcrc in tlic following amnunts (volume pEr cent): He, 0.00052; Ne, 0.0018; A, 0.93; Kr, 0.0001; Xe, 0.000008. Ifc is produccd by radioactive decay of U and T h but is also lost to outer space. A40 is prnduccd by the radioxtive potassium 40 and is the major isotope of argon in tile atmorplicrc.

Thc argon and Iieliiim contents of rocks will vary with thcir age owing to the effcct of radioactivc decay.

Tlic estimated rarc-gas contents of igneous rocks arc (in cc pcr gm of rock): He, 6 x I@; Nc, 7.7 x 10% A, 2.2 x 10% Kr, 4.2 x IO-9; Xc, 3.1 x 10-10.

C: These clcnirnts do not occur naturally in the 13arth’s crust. D: Tlic data for these elements arc missing or unrcliable. E : All t h c x clemcnts are present 3s radioxtivc nuclides in the decay schcmcs of U and Th. I:: These elements Lccur naturally only as 3 consequence of neutron capture by uranium.

TUREKIAN A N D WEDEPOHL, TABLE 2 Geological Society of Anicrica Bulletin, volume 72

. . . . . . - TUKERIAN AND WEDEDOHL-DISTdIBUTION OF ELEMEN.TS .

Fleischer, M., 1955, Estimates of the abundances of some chemical elements and their reliability, p. 145-154 in Poldervaart, Ark, Editor, Crust of the Earth: Geol. Soc. America Special Paper62, 762 p.

Frohlich, F., in press, Beitrag zur Geochemie des Chroms: Geochim. Cosmochim. Acta

. . . . 188 . . . . -

- Cast, P. W., 1960, Limitations on the composition of the upper mantle: Jour. Geophys. Research, v. 65,

p. 1287-1297 Goel, P. S., Kharkar, D. P., Lal, D., Narsappaya, N., Peters, B., and Yatirajam, V., 1957, The berylliuin-

Goldberg, E. D., and Arrhenius, G.O.S., 1958, Chemistry of Pacific pelagic sediments: Geochim.

Goldschmidt, V. M., 1951, Geochemistry: Osford, Clarendon Press, 730 p. Goldschmidt, \T. hf., and Strock, L. W., 1935, Zur Gcochemie des Selens 11: Nachr. Ges. Wiss. Gottingen

Gordon, M., and Murata, K. J., 1952, Minor elements in Arkansas bauxite: Econ. Geology, v. 47, p. 169- I79

Gottfried, D.. \Varing, C. L., and \\.'orthing, H. W., 1956, Hafnium content, hafnium to zirconiuni ratio, and radioactivity of zircon from igneous rocks (.lbstract): Geol. SOC. America Bull., v. 67, p. 1700

Green, I., 1959, Gcoclicmic:il table of the elements for 1959: Geol. Soc. :\ineric:i Bull., v. 70, p: I 1 1;- 11%

Green. I.. and hlilcrv:i:irt, ;\., 1955, Some basaltic provinces: Gcochim. Cosmochiin. :\cta, v. 7, 1). 177- I SS

(;riiii:ilili. I:. S.. 1960. Determination of n i c h i u n i i n t h e parts per million r:ingc in rocks: .An:iI. Clicm.,

I l:ini:igiiclii, II., :in11 Kurod;i. It., 1959, Silver content ol' igiicoiis rocks: Gcocliini. Cosinochiin. :\st:!.

l~l:ini:iguchi, 11.. Reed, G. \\'., :inJ 'I'urkcvicli. ;\., 1957, Cr:iniuiii and barium in s tow nictcoritcs:

H:irdcr. H., I959:i. 13citrag zur Gcoclicniic tlcs I h s . I. Htrr i n X~lincr:rlcn und h.l:igni:itisclicn Gcstciiicn

__ 19591~. Rcitr:IS zur Gcoshcinic tlcs 13ors. 11. Ih)r in Scdmcntcn: ; \ k x l . \ \ ' i s . Giittiiigcn, 11. Sf:i [ l i . -

I-lcidc. I:.. and 138liiii. G.. 1957, Gcoclicmic des Quccksilbcrs: Chemic d. Erdc. v. 19, 1'. 1%-101 I-lciitc. I:.. :ind I-crz. 1-1.. 1355. Zur Ccoclicmie des Blcics: Clieniic d. Erdc. v. 17, p. 217-222 I-Iciilc. F.. and Sinscr. E., 1950. Zur Gcochcmic tlcs Cu und Zn: Der X:iturn.iss.. v. 37, p. 511-542 I - l in t . D. 1.1.. :ind Nicliolls. G. 11.. 1955. Tcchniqucs i n scdiincntary geochemistry: ( I ) Scpar:ition of the

I-lolscr. \V. T.. \\:arncr. I,. :\., \\.'ilinarth. \:. R.. and Cmeron. E. N.. 1951. Notes on tlic geochcinisrry

Holyk. \V.. :ind :\hens. I,. 13.. 1953. Potassium in ultramafic rocks: Geochim. Cosmochiin. ;\ctn. \.. 1.

I-lorstman. E. I-.. 1957. The distribution of lithium. rubidium and caesium i n igncous :ind scdinicntnry

I-liigi. T.. 1956. \'ergleichcntlc pctrologischc iind gcochcinischc Untersuchungcn an Graniten des ;\arm:ls-

Hutchinson. G. E.. Benoit. R. J., Cottcr. I\:. B.. and Wangersky, P. I.. 1955. O n the nickel. colxdt. nnd

10 concentration in deep-sea sediments: Deep-sea Research, v. 4, p. 202-210

Cosmochim. Acta, v. 13, p. 183-212

.Liath.-Phys. KI. IV., V. I , p. 123-143

v. 3 2 , p. 119-111

v. 17. p. -11-52

<.;cochiiii. Ccmiochiin. :\cta. v. 12, 1). 337-317

S a c h r . : : \kid. \ \ : i s . (;iittiiigcn, 11. M:i t l i . -p l iy . K1.. v. 5 p. 67-122

p h ~ s . Kl.. v. 6 . 11. 123-1 83

clctrital :uid non-iletrit:il fractions of liincstoncs: lour. Sed. Petrology. v. 28. p. 465-181

0 1 berylliuin (:\bstr:ict): Gcol. Soc. :\merica Bull.. v. 62, p. 1150-lijl

[I. 1i1-250

rocks: Gcochim. Cosmochim. :\cta. v. 12, p. 1-23

sivs: Beitr. Geol. karte Scli\veiz., Ncuc Folge 91, 56 p.

copper contents of dccp-sea sediments: Nat. :\cad. Sci. Proc., v. 41. p. 160-162 Ishiinori. T., 1951. Gehalt an Molybdin i n jnpanischen vulkanischen Grsteinen: Bull. Cheni. Soc. Japan,

v. 25, p. 251-252 Ishimori. T.. and T:ikashima, Y., 1955, Thallium content of Japanese igneous rocks: Kyusliu Univ. Faculty

Jelfcry, P. G.. 1959. The gcochemistry of tungsten. with special reference to the rocks of thc Cg:lnd:i Sci. Slem.. ser. C.. Chcmistry, v. 2, p. 65-71

I'rotectornre: Geochim. Cosmochim. Acta. v. 16. p. 278-295 Josr. K.. 1932. fiber den \'anacliumgehalt der Scdiinenrgcstcine und sedirnentiircn I.:lgcrstittell: Clicmic

(I. Erde. \.. 7, p. 177-290 K i y q 11.. 1951. North ;\incric:in geosynclines: Geol. Soc. .\meric:i h4cm. 1s. 113 p.

I !

REFERENCES CITED 189

Kokubu, N., 1956, Fluorine in rocks: Kyushu Univ., Faculty Sci. Mem., ser. C., Chemistry, v. 2, p. 95- . . 149 Koritnig, S., 1951, Ein Beitrag zur Geochemie des Fluor: Geochim. Cosmochim. Acta, v. 1, p. 89-116 . Kosterin, A. V., Zuev, V. N., and Shevaleevaskii, I. D., 1958, Zr/Hf ratio in zircons in some igneous

rocks of Northern Kirgizia: Geochemistry (USSR), English ed. (Geochemical Society) no. 1, p.

Kuroda, P. K., and Sandell, E. B., 1954, Geochemistry of molybdenum: Geochim. Cosmochim.. Acta,

. 116-1 19

V. 6, p. 35-63 Kuroda, R., 1957, Geochemical investigations of granite: Jour. Chem. SOC. Japan, Pure Chem. Sec.,

v. 78, p. 141 Liebenberg, C. J., 1956, The spectrochemical determination of Cs in igneous rocks, using a chemical

Macpherson, H. G., 1958, A chemical and petrographic study of Pre-Cambrian sediments: Geochim.

Mason, B., 1958, Principles of geochemistry: 2d ed., New York, John Wiley, 310 p. Masuda, A., 1957, Simple regularity in the variation of relative abundances of rare earth elements: Jour.

Earth Sciences, Nagoya Univ., v. 5, p. 125-134 Merrill, J. R., Honda, M., and Arnold, J. R., 1958, Beryllium geochemistry and beryllium-10 age de-

tefmination: Proc. 2d United Nations Internat. Conf. Peaceful Uses of Atomic Energy, Geneva,

Merrill. J. R., Lyden, E. F. X., Honda, M., and Arnold, J. R., 1960, The sedimentary geochemistry of

Minami, E., 19353, Selen-Gehalte von europaischen und japanischen Tonschiefern: Nacher. Ges. \Viss.

- 1935b, Gehalte an selcenen Erden in europiischen und japanischen Tonschiefern: Nadir. Ges. \Viss.

Morita, Y., 1955, Distribution of copper and zinc in various phases of the earth materials: Jour. Earth

Mullin, J. B., and Riley, J. P., 1956, The occurrence of cadmium in s e a ~ a t e r and in marine organisms

Murray. E. G., and Adams, J. A. S.. 1958, Thorium, uranium and potassium in some sandstones: Geocliim.

Nockolds, S. R., 1954, Average chemical composition of some igneous rocks: Geol. Sm. America Bull.,

Nockolds, S . R., and .Allen, R., 1953, The geochemistry of some igneous rock series. Part I: Gemhim.

concentration procedure: Geochim. Cosmochim. Acta, v. 10, p. 196


V. 2, p. 251-254

the beryllium isotopes: Geochim. Cosmochim. Acta, v. IS, p. 108-129

Goctingen, Math.-phys. KI. IV, v. I , p. 143-145

Gottingen, Math.-phys. KI. IV, v. 1, p. 155-170

Sciences, Nagoya Univ., v. 3, p. 33

and sediments: Jour. Marine Research, v. 15, p. 103-122


v. 65, p. 1007-1032

Cosmochim. .Acta, v. 4. p. 105-142 - 1956, The geochemistry of some igneous rock series. Part 111: Geochim. Cosmochim. Acta, v. 9,

p. 34-77 Okada, S., 1955, Chemical composition of Japanese granitic rocks in regard to petrographic provinces.

- 1956, Chemical composition of Japanese granitic rocks in regard to petrographic provinces. Part IV:

Onishi, H., 1956, Geochemistry of germaniuni: Bull. Chem. SOC. Japan, v. 29, p. 686-694 Onishi. H., and Sandell, E. B., 19553, Geochemistry of arsenic: Geochirn. Cosmochim. .Acta, v. 7, p. 1-33 - 1955b, Notes on the geochemistry of antimony: Geochim. Cosmochim. Acta. v. S, p. 213-221 - 1947, Meteoritic and terrestrial abundance of tin: Geochim. Cosmochim. Acta. v. 12, p. 262-270 Ostrom, XI. E., 1957, Trace elements in Illinois Pennsylvanian limestones: Illinois State Geol. Surv.

Ottemann, J. , 1940. Untersuchungen von Spurenelementen, insbesondere Zinn. in Tiefengesteinen und

Petterson, H., and Rotschi, H., 1952, The nickel content of deep sea deposits: Geochim. Cosmochim.

Picciotto, E., and \\'ilgain, S., 1954, Thorium determination in deep-sea sediments: Nature, v. 173,

Pinson, W. H., Ahrens, L. H., 2nd Franck, A t . L., 1953, The abundances of Li. Sc. Sr. Bn and Zr in

Part 111: Science Rep. Tokyo Kyoiko Daigacu, Sec. C, v. 4, p. 163-184

Science Rep. Tokyo Kyoiko Dsigacu, Sec. C., v. 5, p. 25

Circ. 243, 34p.

einigen gesteinsbildenden Mineralien des Harzes: Z. angew. Mineralogie, v. 3, p. 142-1 69

Acta, v. 2, p. SI-90

p. 632-633

chondrites and ultrabnsic rocks: Geochim. Comochim. .Acta, v. 4, p. 251-260

. . . . 1 .

.: 190 .TUREKIAN AND- WEDEPOHL-DISTRIBUTION OF ELEMENTS

Preuss, E., 1940, Beitrige zur spektralanalytischen Methodik. 11. Bestimmung von Zn, Cd, Hg, In, TI, Gc, Sn, Pb, Sb und Bi durch fraktionierte Destillation: Z. angew., Mineralogie, v. 3, p. 8-20

Rankama, K., 1944, O n the geochemistry of tantalum: Bull. de Comm. geol. Finlande no. 133, 78 p. - 1948, O n the geochemistry of niobium: Ann. Acad. Scient. Fennicae A 111, no. 13, 57 p. Rankama, K., and Sahama, Th., 1950, Geochemistry: Chicago, Univ. of Chicago Press, 912 p. Reed, G. W., Kigoshi, K., and Turkevich, A., 1958, Activation analysis for heavy elements in stone

meteorites: Proc. 2d United Nations Internat. Conf. Peaceful Uses of .Atomic Energy, Geneva, v. 28, p. 486-490

Ricke, W., in press, Ein Beitrag zur Geochemie des Schwefels: Geochim. Cosmochim. Acta ROSS, C. S., Foster, M. D., and Myers, A. T., 1954, Origin of dunites and of olivine-rich inclusions in

Runnels, R. T., and Schleicher, J. A., 1956, Chemical composition of Eastern Kansas limestones: State

Sahama, T. G., 1945, Spurenelemente der Gesteine im siidlichen Finnixh-Lappland: Bull. de Comm.

Sandell, E. B., 1946, Abundance of tungsten in igneous rocks: Am. Jour. Sci., v. 244, p. 643-648

- 1952, The beryllium content of igneous rocks: Geochim. Cosmochim. :\cta, v. 2. p. 211-2115 Sandell. E. B., and Goldich, S. S., 1943, The rarer mctnllic constituents of some American igneous rocks:

Sha\v, D. IM., 19523. The geochemistry of thallium: Geochim. Cosniochim. Acta, v. 2, 11. 118-154 - 1952b, The geochemistry of indium: Gcocliim. Cosmochim. k t a . v. 2. p. 185-206 - 1954, Trnce elements in pclitic rocks I, 11: Geol. SOC. America Bull.. v. 65, p. 1151-1166, 1167-1152 - 1957, Some aspects o l the detcrrnination of bnriuin in silcatc rocks: Spcctrochim. Acta, v. IO, p.

Shepherd. E. S., 1940, Note on the fluorine content of rocks nnd occan bottom snmplcs: ;\in. Jour. Sci.,

Smales. :\. A., 1955, Some trace-element Jcteriiiinations in G. I a n d \\'-I by neutron 3ctiv;itioii: Geocliim. Cosmochim. Acta, v. 8, p. 300

Smales, A. A., and Salmon, L., 1955. Ilcterniinntion by rndionctivation of smdl nmouiits of urbidiuni and caesium in sea-water and related matcri:ils of geoclicmical interest: h a l y s t . v. SO, p. 37-50

S d c s , A. A., Mapper. D., and Wood, A. J., 1957, Thc determination. by radioactivation. of smnll quantities of nickel, cobalt, and copper in rocks, innrinc sediments and mctcoritcs: :\nnlyst, v. 82, p. 7 5 4 3

Starik, I. E., Kuznetsov, Y . V., Grashchenko. S. hf., and Trenklikh, \ f . S.. 1958. The ionium method of determination of marine sediments: Geochemistry (CSSR), English ed. (Geochemical Society), no. 1, p. 1-15

basaltic rocks: Am. Mineralogist, v. 39, p. 693-737

Geol. Survey Kansas Bull. 119, pt. 3, p. 81-103

Geol. Finlande no. 135

. - 1949, The Ga content of igneous rocks: Am. Jour. Sci., v. 247, p. 4 0 4

I, 11: Jour. Geology, v. 51, p. 99-115, 167-189

125-127

V. 238, p. 117-128

Stock, A., and Cucucl, F., 1934, Die Verbrcitung des Quccksilbers: Naturniss., v. 22-24. p. 390-393 Strock, L. W., 1936, Zur Geochemie des Lithiums: Nachr. Ges. \Viss. Giittingcn .\lath.-Phys. KI. IV,

V. 1, p. 171-204 Sugawara. K., and llorita, Y., 1950, On the revision of the Clarkc numbers of copper and zinc: Mik-

rochemie, v. 36/37, p. 1093-1099 Tatsumoto, M., 1957, Chemical investigations of deep sea deposits (nos. 21-26): Ni, Co, Cu. Zn, Sn, Pb.

Be content of deep sea deposits: Jour. Chem. SOC. Japan, v. 77, p. 1637, v. 78, p. 38, 42, 405, 409, 502 (in Japanese)

Tauson, L. W., and Pevtsova, L. A., 1955, On the relationship of the distribution of lead and zinc in the rocks of the Caledonian granite complex of Susamyr (Central-Tienshan) : Doklady )\cad. Nauk,

Tilton, G. R., and Reed, G. W., 1960, Concentration of lead in ultramafic rocks by neutron activation

Tourtelot. H. :\., 1957. Chemicnl composition of the Picrrc shale and cquivnlent rocks of Late Crctaccous

1 roger. E., 1934, Der Gehnlt nn seltcnercn Elementcn bei Eruptivgcsteinen: C%mie tl. I<&, v . 3, 13.

USSR, V. 103, p. 1069-1072

analysis (Abstract): Jour. Ceophys. Research, v. 65, p. 2529

age, Great Plains region (Abstract): Ccol. SOC. America Bull., v. 68, p. IS06 ,. 286-3 IO

. . . . . . ‘REFERENCES CITED 191 .

. .

. .

Turekian, K. K., 1956, The abundance of Cu, Ni, and Cr in basaltic rocks (Abstract): Am. Geophys.

- 1960, Elements (geochemical distribution), in Encyclopedia of Science and Technology, v. 4: New .

Turekian, K. K., and Armstrong, R. L., in press, Magnesium, strontium, and barium concentrations and

Turekian, K. K., and Carr, M. H., 1960, The geochemistries of chromium, cobalt, and nickel: XXI Inter-. _:I

Turekian, K. K. and Carr, M. H., in press, Chromium, cobalt and strontium in some Bureau of Standards

Turekian, K. K., and Feely, H. W., 1956, Variations in the abundance of trace elements in a eupelagic

Turekian, K. K., and Kulp, J. L., 1956, The geochemistry of strontium: Geochim. Cosmochim. Acta,

Vincent, E. A., and Crocket, J. H., 1960, Studies in the geochemistry of gold-11: Geochini. Cosmochim.

Vincent, E. A, and Smales, A . A., 1956, The determination of palladium and gold in igneous rocks by

Vinogradov, A. P., 1956, The regularity of distribution of chemical elcments i n the .Earth’s crust:

Vinogradov, A. P., and Ronov, A. B., 1956, Composition or scdiinentary rocks ol the Russian pl:itform

\’inogr:idov, A . P., Vainslitcin, E. E., and Puvlenko, L. I., 1955, Tungsten and molybdenum in igneous

\V:igcr, L. R.. Smit, J. van R., and Irving, H., 1923, Indium content of rocks :ind minerals from rhc

\Vnsscrstcin, 13.. 1951, South African granitcs and their boron contcnt: Geochim. Cosmocliim. Acta, v. I ,

Weber, J. N., 1960, Gcochcmistry of graywackes and shalcs: Scicncc, v. 131, p. GG-1-665 \Vcdcpohl. K. I-I., 1953, Untcrsuchungcn zur Gcociicmic dcs Zinks: Gcochiin. Cosniochiin. :\ct:i, v. 3,

__ I 95 5, Schivcr me tal Is e ha I t c tlcr K;ilkgcrus t c ci n iger iiiw incr ( ) r p n isiiic 11 : S;ic h r . ;\ k:ic I. \ V i s .

- 1956, Untersuchungcn zur Geochemic dcs Blcis: Gcochim. Cosmochim. Acta. v. IO. 11. G9-liS _- 1960, Spurenanalytische Uiitcrsuchungen a n Tiefsectoncn nus clcin :\tlantik: Geochim. Cosmochim.

Whitfield, J. A I . , Rogers, J. J. W., and /\dams, J. A. S., 1959, The relationship between the petrology and the thorium and uranium contents of some granitic rocks: Geochim. Cosmochim. Acta, v. 17, p. 24s-271

Znamensky, E. E.. 1957, On the average contents of niobium and tantalum in igneous rocks and the crust of the earth: Geochemistry (USSR), No. 8, p. 730

. Union Trans., v. 37, p. 361

York, McCraw-Hill Book Co., p. 549-552

calcite-aragonite ratios of some recent molluscan shells: Jour. Marine Res.

nat. Geol. Congress, Copenhagen, Part I, p. 14-26.

rock reference samples: Geochim. Cosmochim. Acta

Atlantic core (Abstract): Geol. SOC. America Bull., v. 67, p. 1739

V. 10, p. 245-296

Acta, V. IS, p. 143-148

radiollctivation analysis: Geochim. Cosmochim. Acta, v. 9, p. 154-160

Geokhimipa, p. 1-52 (A.E.R.E. Lib/Trans. 795; 1957)

in connection \vitli thc Iiistory of its tectonic movements: Geocheniistry (USSR), no. 6, p. 3 - 2 ) .

rocks: Geochemistry (USSR), English cd. (Geochemical Society), no. 5, p. 497-509

Skxrganrd intrustion. East Greenland: Geochim. Cosmochim. Acta, v. 13, p. SI-SG

p. 329-338

11. 93-1i2

Gottingcn >Iath.-I’hys. KI. 5, p. 79

>\eta, V. IS, p. 200-231

i b f m u s c n r i ~ r RECEIVED n Y T I I E SECRETARY of: T I I E SOCIETY, J A N U A R Y 21, 1959

. . . . . .

. . .. . . . . . .

. . . .

P A P E R S I N P R E S S F O R F O R T H C O M I N G I S S U E S i

WELDED TUFFS A N D FLOWS I N THE RHYOLITE PLATEAU OF YELLOWSTONE PARK, WYoairNc. Francis R.

PRECAMBRIAN ROCKS A N D LARAMIDE STRUCTURE ALONG THE EAST FLANK OF THE BIGHORN MOUNTAINS

LIMNOLOGY, SEDIMESTATION, A N D MICROORGANISMS OF THE SALTON SEA, CALIFORNIA. Robert E. Arnal

AGE A N D STRATIGRAPHIC RELATIOSS OF THE FORMOSA REEF LIMESTONE OF SOUTHWESTERN ONTARIO,

HYDROTHERMAL MAGNETITE. William T. Holser and Cecil J. Schneer

PALEOTEMPERATURE ANALYSIS OF THE PLIO-PLEISTOCENE SECTION AT LE CASTELLA, CAL.*BRI.\, SOUTHERN

Boyd

NEAR BUFFALO, WYOLIINC. Richard A. Hoppin

CANAD.$. J. A. Fagerstrom

ITALY. C. Emiliani, T. Mayeda, and R. Selli

N E W F 0 R E I G N B I B L I O G R t\ P H Y

BIRLIOCR.*.PHY . \SD INDEX OF GEOLOGY EXCLUSIVE OF NoRrii :\MERICA. Volume 23. Starie Siegrist,

Price, $13.00 Special price to Fcllocvs and h k m b c r s of GSA, $3.50

Mary C. Crier, and others.

N E W h.1 E M 0 I R

SZ. OLIGOCESE PL.\s.rs I:ROM ~ i i i i UITKR R ~ B Y RIVER D.\sIs, Soti.rii\vT:smits lfos.r.\s.\. Herman F. I3 cc k cr

Description o f ;i cvcll-preserved h a (37 families, GI genera. and 81 identifiable species, 25 of which arc ncn.) fro111 Tertiary lossililcrous shales of the Ruby K i w r b:isin. The author uses the flora to intcrprct the palcoccology ;ind paleoclimatology and suggcsts that the Ruby and Florissant as- scmbl:iges represent n single botanical province.

125 pages, i figures. 32 photographic plates . Price, $4.75 Special price to Fellocvs and l4crnbers of GS:\, $3.00

M E h 4 O I R I N P R E S S

s j . BIOSTRhTlGRAI’HIC STCDIES IN TIIE C O X I A S C H E (CRETACEOUS) SERIES OF . JRTllERN ?VlEXlCO A N D TES.\S. Bob F. Perkins

T h e author relates the biostratigraphy of the :\urora limestone of southwestern Coahuila to the type :\urora and to the type Loner Cretaceous of northern Texas and reports stratigraphic revisions of the Fredericksburg and Washita groups of north Texas. H e describes a fauna of larger inverte- brates from the upper member of the tlurora and on paleontological evidence correlates the member cvith the Fort Worth liniestone and Grayson marl. 1 - i ~ concludes that the :\urora limestone is of cpiiicritic origin, deposited far from any source of terrigenous clastic material.

;\bout 1-10 pages, 19 figures, 3 maps in color, 29 photographic plates Espected date of issue February 25 Pre-publication price $ 5 . 2 5 ; post-publication price $6.00. Special price to Fellows and l iembers of GS:\ S4.50

of...sea sediments. “ igneous’’ rocks under this heading we include some ultra- basic rocks...

Documents