Chemical exchange between sea water and deep ocean basalts

EARTH AND PLANETARY SCIENCE LETTERS 9 (1970) 269-279 NORTH-HOLLAND PUBLISHING COMP

CHEMICAL

EXCHANGE

BETWEEN SEA WATER AND DEEP OCEAN BASALTS Roger HART U s Naval Oceanographw Offtce Prolect GOFAR, Washington D C, U S A Recewed 29 June 1970

Analysis of 112 deep ocean basalts reported m the hterature show systematic chemical trends m major elements with distance from ridge spreading centers These same trends are observed m the weathenng of single ptUows and produce a chemical composlUon similar to volcamc alkah basalts. Upon hydration by sea water, tholentle ridge basalts gwe off sd~ca, calcmm, magnesmm, and gain potassmm, tron, tltamum, manganese, sodmm and phosphorus Aluminum shows no apparent change. The yearly contnbutaon to sea water i n 10-9 grams of the major elements per cm 3 of basalt Is esttrnated at S1=+7 4, Tl=-I 6, Fe=-3 3, Mn=0 06, Mg=+2.3, Na=-0 21, Ca=+2 1, K=-0 9, and P=-0.14

1. Introduction The concept that the major element chemistry o f basalt rocks varies systemahcally with distance from the major oceamc ridge was first documented by McBlrney and Gass [ 1] who showed that basalts from volcanic oceamc ~slands decrease in sthca saturahon with distance from the ridges They correlated the decreasing slhca content with dtmmlshmg heat flow away from the ridge Engel and Engel [2] noted that basalts dredged from the ridge axis are unlformally tholenhc m nature while samples taken on seamounts at a distance from the ridge are more hkely to be alkali basalts. They attributed the phenomenon to a process o f crystal differentiation in the magma chambers o f sea floor volcanoes, however, since the work o f Yoder and Tllley [3], crystal dffferentmtlon is no longer considered a viable mechanism for producing an alkali basalt from a tholeutlc magma In order to explain chemical vanahon of basalt samples from the Confederation Peak regton o f the Mid-Atlantic Ridge, Aumento [4] suggested a process of progressive degrees o f parhal melting o f a pyrollhc melt following from the work of Green and Rmgwood [5]. The concept o f decreasing partial melting was correlated to progresswely greater depths with distance from the

ridges by Kay, Hubbard, and Gast [6] and generalized to apply to all major ridge systems With the general acceptance o f the sea floor spreadmg theory o f Hess [7] and Dletz [8] and the magnetic stratigraphy theory o f Vine and Mathews [9], it is now possible to suggest that systematical chemical trends correlatable with distance from ridge axis are also related to increasing age o f the sea floor that has moved out from the ridge spreading centers. Simultaneously with observahons o f varmtlons m magma types across the oceamc ridges, various authors have noted the effects o f sea water woathenng on basalts extruded on the sea floor. One o f the most important and consistent observatxons made is that as alteration proceeds the water content o f the sample increases due to the formation o f hydrated alteration minerals. Engel et al. [10] reported that alterahon of ridge basalts causes an increase in H20, and Fe203 contents of basalts and a loss m MgO and total iron. Nichols [ 11 ] found a decrease in the percentage o f calcmm, magnesium and sodmm and attributed the losses to leachmg b y sea water. Moore [12] showed that palagonlzatlon o f basalt glass resulted m a loss of CaO, a lesser loss o f Na, and enrichment m K and Fe. Hart and Nawalk [13] reported that the CaO and S10 2 contents o f basalts from the Puerto R~co trench decrease with increasing water content Hart [14]

270

R.HART

showed that K, Rb and Cs are enriched m altered portions of ridge basalts by factors of 2, 5, and 20 respectively Myasbaro et al [ 15] showed that the Fe203 content of basalts increased with the H20 content. There seems to be some agreement among the authors that the Fe203 and K contents of basalt are enriched whde the CaO content is depleted upon weathenng There is disagreement among the authors as to whether total iron goes up or down The loss of Na, Mg, and $102 has been suggested but has not yet been confirmed or contested by the above ment~oned authors A more recent approach to the problem of sea water weathering of ocean floor basalts has been to study the concentric zones of alteration visible within single boulders and pillows. There is stdl some disagreement as to whether the zones are produced by sea water weathering after the magma has cooled or produced by deuterlc alteration during the magmatlc stage Myashlro et al. [15] reported three zones of visible weathering m a single basalt boulder The outsxde rim is weathered to a yellow brown zone that grades into a transitional grey-brown zone surroundmg the unaltered dark grey core They show that the outer weathered rim is enriched in MnO, T102, Fe203, P205 and total Iron whde depleted m $102, FeO, MgO and CaO relative to the core Myashlro's analysis shows no conclusive trends for A1203, Na20, or K Paster [16] recognized three distinct zones of alteration within lndwldual basalt pdlows dredged from the South Pacific 1) an outer zone of hydration depleted in magnesium and calcmm and enriched m total iron and potassmm, 2) a zone of serpentlnlzatlon adjacent to joints and fractures attributed to deuterlc alteration between sea water and the coohng lava body, 3) a zone of chlontlzatlon m the core of the sample with depletion of magnesium and total ~ron attributed to high temperature deuterlc alteration Paster [ 16] suggested that chemical alteration of basalt m~ght have a major impact on the geochemical budget of the major elements m sea-water. He particularly emphasized the enrichment of sea water m iron and manganese in an effort to prowde a source for manganese nodules His results also show alteration of basalts enriches sea water m s~hca, calcium, and magnesmm. Paster's [16] emphasis on deutenc alteration is m opposmon to the interpretations of Myash~ro et al

[15] and Hart [14] who show that the alteration 1s most intense in the outer zones rather than the mner zones and therefore most hkely due to sea water weathering as opposed to deuterlc alteration The tmportance of deuterlc alteration cannot be dismissed, particularly m thick lava bo&es, but in small pillows the weathering effects of Mlyashlro et al. [ 14] and Hart [ 13] are probably superimposed over the deuterlc effects particularly m samples that have been in contact with sea water for some time Very httle work has been done in the petrology of the weathering processing. In adchtlon to Paster's [16] observations of chlontlzatlon and serpentmlzatlon, Nawalk [ 17] reported the replacement of plagloclase by smectlte and senclte Nawalk [17] feels that at least part of the alteration he observed is deutellc in nature Engel and Engel [18] reported olivine altered to lddmgslte and hrnonlte and reported the presence of chlorite and iron oxides.

2. Results Fig. 1 shows a plot of the percent potassium content versus age for 112 analyses of ridge basalt. The sample location and hterature source are ln&cated for each analysis. The magnetic reversal ages were determined by plotting sample locations on maps of ocean floor residual magnetic field reversals of Vogt et al. [32], Hlertzler et al [33] and Hays and Pitman [34]. Potassmm argon ages were assigned to the reversals back to 4 mybp by Cox et al [35] and extended to 80 mybp by paleontological dating of the oldest sednnent in JOIDES cores by Maxwell [36] Some samples could not be assigned ages because of lack of magnetic data and were not included m thas study The estimated age uncertainty for most samples is + I my. Samples with a large number of plagloclase or ohvme phenocrysts were not included m order to exclude crystal &fferentlatlon effects. In an effort to avoid confusing true alkali basalts with weathering effects, alkahc samples with less than 1% H20 were judged to be fresh alkah basalts, and not included in this study The most striking feature of fig. 1 is the systematic increase in the scatter of potassium values w~th age. Another important feature is that the medmn potas-

CHEMICAL EXCHANGE BETWEEN SEA WATER AND DEEP OCEAN BASALTS

1.8

271

[] RIDGE BASALT • WEATHERED RIDGE BASALT

m16

~

1.6

INCLUDES

1.4

'~5 ! 3 I 7 $ I

m_ml4

1o2

"19 m14

MID-ATLAWrlC RIDGE 'M*--30*N MA RIDGE 29"N MA RIDGE 45"N MID-INDIAN OCEAN RIDGE GORDA RIDGE MID-INDIAN OCEAN RIDGE EAST PACIFIC RISE

toO 1 MA RIDGE 24"--30"N

0.8

2 3 4 5 6 7 8 9 I0 11 12 13 14 15 16 17 18 19

a2 BX4

0.6

1115

03 o3

o2 02 ~2

114

n3

o2

o18 m14 1114

0.4 n3

0.2

o.%

13 mX

~ X l °7 kl~2 n18 ~11 R77

o13

o16

ol

lb

;2

is

;o

M A RIDGE 29°N MA RIDGE 4S*N MID-INDIAN OCEAN RIDGE GORDA RIDGE MID-INDIAN OCEAN RIDGE EAST PACWIC RISE MA RIDGE 22°N EAST PACIFIC RISE M A RIDGE 30*N MA RIDGE 22*N--52*N M A RIDGE 10*N--30*S CARLESBURG RIDGE BALD MOUNTAIN NORTHEAST PACIFIC HENDERSON SEAMOUNT JUAN de FUCA RIDGE MID-ATLANTIC RIDGE 28"--30"N WEST COCOS SEAMOUNT

20

&GE ~MAGN~'T|G REVERSAl=) I06YEARS Fig 1 A plot of percent potassmm versus residual magnettc field reversal age for 112 analyses of ridge basalts The ages of the samples were determined by plotting the sample locations on residual magnetic field reversal maps ofVogt et al [32] for Northern Atlantw, Htertzler et al [33] for Indian Ocean and Southern Pacific, and Hays and Pitman [ 34] for Northern Pacific The general sample locations for each analysts are mdtcated and taken from the foUowmg hterature sources 1) Mwashtro, Shldo and Ewmg [ 15 ], 2) Mutt and Tllley [ 19], 3) Aumento [ 20], 4) Engel, Fisher and Engel [ 10], 5) Kay, Hubbard and Gast [6], 6) Engel and Fisher [31], 7) Bonattl [22], 8) Melson, Thompson and Van Andel [23], 9) Engel and Engel [24], 10) Mmr and Tdley [25], 11) Ntehols, Nawalk and Hays [26], 12) Engel and Engel [27], 13) Cann and Vine [27], 14) Aumento and Loncavenc [29], 15) Engel and Engel [ 18], 16) Poldervaart and Green [30]% 17) Kay, Hubbard and Gast [6], 18) Kay, Hubbard and Gast [6], 19) Engel and Chase [ 31 ]. smm value o f ridge basalts appears to increase with age. Two interpretations o f t i m plot immediately come to mind The first interpretation follows from the suggestions of McBlrney and Gass [1 ] , Aumento [4], and Hubbard, Kay and Gast [6] that basalt magma changes Its chemical composition with distance from the ridges and that a plot such as fig. 1 indicates a d d m o n s from a single progresswely changing magma source, or from numerous sources o f a different nature. The other interpretation is that fig 1 is the result o f

samples progressively altered with time and that the htgh potassium values represent samples with a higher percent o f altered material, whde the low potassmm values come from samples or sections o f samples such as the center portions whach are relatwely fresh. In order to emphasxze the importance of weathering in producmg the plot shown in fig. 1, samples that were identified as altered by the authors are shaded. The means and standard deviation for all twelve major elements, plus total iron are shown in fig 2. The plots come out remarkably close to straight hnes for

R HART

272 4O

~0

2

leo.

3

4110 SO"

II

mio 44.O"

io. 4ZO"

O~

~G

eo

t-o

4~

tO

FERROUS

OI

FERRIC I R O N oo ~

O0

loft

7

~*

ItO

9

~0" It

CALCIUM 40

lid

il PHOSPHORUS 14

.0I

12

I0 O4

•0

O!

~0

moe

TOTAL IRON AS FERROUS IRON

02

OG

oo AG[ iN M v

AG[ IN uY

AGt it* t r y

0

Fig. 2. Mean and standard dev]atlon plots of oxides of the twelve major elements versus residual magnenc field reversal age

CHEMICAL EXCHANGE BETWEEN SEA WATER AND DEEP OCEAN BASALTS

273

6 O0

o%~

50(:

~ •

~

8o

%%

400 ~

\

o

~\

'" W " " . . %

o/ io°O ° o . . . . 0 . m

• ---

idvm+ ( m ~ .

/

ore,s)

x

5 m IAV~°IF+**Jl n.~+l Umll

X%o//

R,~. moll ~ 1 ~ ~ m l h m co~ lor,m boe~ +rune 1 ~ l d y o l ~ , . o (1~9) C~+I Iv+rlm pill0511 t r e n d f + P o s t e r (19(181

Cw+~,,m a.llo. IrlmeImml'~ller ~ l ~ l

IOO

o

4200

4400

/

,1~OO ~s,o~

4800

JOOO

J8OO

Fig 3 A plot of the total alkalis versus mhca for ndge basalts and core-to-nm weathering trends The ridge basalts axe contoured according to age to show how progressnve weathering of a tholentic basalt produces a basalt of alkah nature most major elements for 18 m y The oxides o f silicon, magnesium, and calcium, show good trends decreasing w~th age; the increase in standard dewations demonstrate a mmllar increase in scatter as shown in fig. 1 for potassmm The slopes o f the lines are relatively steep compared to the standard deviation The oxides o f titanium, potassium and phosphorus, plus total iron as FeO amd water all increase with an expanding standard deviation fields and mean slopes that are significant. The oxides o f aluminum, manganese and sodmm do not show significant trends, although a case can be made for small increases in manganese and sodmm. Ferric iron replaces ferrous iron as would be expected during the weathering process. The increase m H20 is also lndlcatwe of weathering All the trends shown in fig 2 can be interpreted as weathering trends when one makes a comparison o f fresh ridge basalts with weathered ridge basalts as m table 1. The weathered ridge basalts in table 1 are

typified b y concentric zones o f brown discoloration, and alteration minerals such as chlorite, lddlngslte and smechte replacing primary olovlne, chnopyroxene and plagloclase An average alkali basalt c o m p o s i n o n is given in table 1 to emphasize that a weathered ridge basalt can be confused with an alkali basalt with the exception o f the low water and oxidized iron contents found m fresh alkali basalt The single sample core to rim weathenng trends of Mlyashlro et al [15] andPaster [16] can be correlated with the ridge basalt age trends o f fig 2. This is done, first o f all, in a plot o f total alkalies versus silica in fig 3 The single sample trends o f M~yashlro et al. [ 15] and Paster [ 16] are plotted along with all the analysis o f ridge basalts used previously to plot fig 1 The ridge basalts are contoured according to age and follow the core to rtm weathering trends. The dividing line between alkali and tholeilhC basalts first used by McDonald and Katsura [35] but now o f questionable

M

~:'--.. .~...

t

4,

8.

/0

,

...,,.//;/

.~.. ",,\

46"

~.

!t

" ,X~

~

SILICA

",,,.

COle lk; r l ~ pillow t m114 from P 4 M ~ (1948)

---0 HYIP (Ndip Ida) - - - - - - I - 4 H Y I P (Cmotal pfOvlll(:e) ..... 5 - H I IdYIP ( F k m i m ) C o ~ Io rim b ~ M e r h, I 1 4 from MtmIIIlim ( l O I I )

4 ¸

:.~~,~

\

\FERROUS IRON

E____

I.~.....:.~ ~/

u

•

o/

ol.

03'

04"

08"

OA'

I0

I$.

/

~\

/

/

/

6 MANGANESE

//

3

/

/

/

._,

//

/

--~-",~\ ALUMINUM

/'~'~- I:,:

/o

,o.

/

;o

t~ -,-.1

ii

/"

-,,

\

POTASSIUM

10

///"

/

/

/ /

MAGNESIUM

7

OC

Ot

~04

06

08

Io

/

,

/ ' I I , I' , I

i/

/

i

~"~

[•':•"""

\

i

I

I '2"

HeO

b

,b

,~,

\ II ~ PHOSPHORUS

"i

\

%%

CALC IUM

8

g

12

I,I

16

o

o

tol

,oil

Cl'Ja.

,"

/

.

....

.

"'\

.

\

I/I

\\

.

.

,

.

SODIUM

FERRIC PLUS FERROUS IRON

.

•

•r

~'Lz

~

i/

~

"°t

~

Fig 4 Plots of the oxides of eleven major elements and ferric plus ferrous iron versus water content for ridge basalts and core-to-rma weathering trends The ridge basalts are contoured according to age and show that the changes m percent chemical composlt]ons are related to water content and known weathering trends

el

•o

r

.1__22

:'~'. I.% > "'-.

>

>

Z

~n

©

m

> Z E7

>

Z

> Z

>

=

276

R HART Table 1 Comparison of fresh ridge basalt with ridge basalt ~dentffted as altered by the authors Chemical constituent

Average of 45 fresh ridge basalt

Average of 24 weathered ridge basalts

$102 TIO2 A1203

49 92 1 53 15 63

47 92 1 84 15 95

Fe203 FeO MnO

1 65 8 19 0 17

4 58 6 05 0 19

MgO CaO Na20

7 65 11 17 2 75

6 38 10 73 2.91

K20 H20 P2Os

0 16 0 95 0 13

0 53 2 19 0 25

Average of 178 alkahc basalts* 46 9 30 15 5 31 86 0.16 6.9 10 4 30 13 08 0 39

* From Manson [40]

usefulness, is included only as a reference to point out that a basalt of apparent alkah nature can be produced from a tholennc basalt by interaction with sea water The core to rLm weathering trends of Mlyashlro [15] and Paster [16] are compared to ridge basalts in a plot of eleven major elements plus total iron versus water in fig 4 In the core to ram trends, the water content is always greatest on the outside emphasizing

Table 2 Basalt weathermg exchange with sea water m 7 X 106 years* Umts = 10-2 g/cm3 Constant AI=0

Lost to sea water $102 = 5 2 MgO --- 2 9 CaO = 2 1 -

Gamed from sea water T102 = 1 1 FeO = 2 6 MnO=005 NaO2= 0 2

-

K20

-

P:~O5 = 0 0 8

-

H20

Total lost 10 2

= 0 9

= 2 6

Total gamed 7 53

* Assuming constant aluminum compos~tlon and a constant specific gravity of 2.60 for ridge basalt.

interchange with sea water as opposed to deutenc alteration. The ridge samples are contoured to age and it can clearly be seen that they pick up water and exchange elements as they get progresswely older. The ridge basalts trends generally show good correlation to the core-rim trends Since the chemical changes reported here are all m percentages it is virtually impossible to estimate absolute gains and losses m the exchange between sea water and basalts. The percentage change of one element may only be an apparent change brought about by the change of a different element An attempt to discuss the real chemical changes going on between sea water and basalts can be made by assuming that aluminum is constant, and that all changes relative to ~t are real This assumption is supported by the common obsewation that aluminum is the most immobile cation during subaerlal weathering of Sdlcate rocks as discussed b y Loughnan [36] By fitting stra.tght hnes to the age plot m fig. 1, an estimate of weight change can be made for each element assuming a specific gravity of 2.60 g/cm3for ocean floor basalt. Fig 2 shows the net gains and losses estimated for a period of 7 X 106 years The net loss is 2.67 X 10 -2 g/cm 3 greater than the net gmn suggestmg that either the density of the rock decreases, the

CHEMICAL EXCHANGE BETWEEN SEA WATER AND DEEP OCEAN BASALTS curves of fig 1 are inaccurate, or the assumption o f constant aluminum is incorrect The loss o f silica and magnesium may be a result o f the alteration o f ohvme and clmopyroxene to chlorite and clay minerals The loss of calcate may come from the alterataon o f anorthlte to smechte which as then converted to dlite extracting potassium from sea water The increase an aron may be a result o f iron oxide d e p o s m o n as oxide coatings and cavaty filhngs Tltamum may go into the f o r m a h o n of dmemte. Potassium probably goes into llllte Phosphorus, manganese and sodmm are probably picked up m small amounts but at is difficult to predict whach alterataon mmerals they would favor. Water is probably added to the rock both an the formation o f clay mmerals and as absorbed lntersntml material. The number o f available analyses is anadequate and one o f the purposes o f this paper as to point out the need for more analysis o f the deep ocean floor, partacularly an regions removed from the spreading centers Furthermore, m future studies o f oceanic rocks, the effects o f weathering must be taken into account. Further studies should also clarify the relationship between deutenc alterataon and weathering In the long range balance o f ocean chemistry, the effects are most hkely addatave. Efforts should be made to dlstmgmsh between true alkah basalt and a weathered ridge basalt of alkali composition, and the underlying reasons explaining two apparently different processes, one magmatac and one weathering, producing rocks of similar composatlon need to be further studied.

3. Geochemical implications The importance o f the chemacal exchange between sea water and basalts an the overall geochemical budget o f the major elements depends largely on the depth to which sea water weathering penetrates the ocean floor Prellmanary results an the JOIDES coring program have revealed altered basalt at a depth o f ten feet Considering the hydrostatic head provided by 4000 m o f sea water and fluxmg abd~ty o f water an mineral reactaons, ~t seems hkely that sea floor basalt weathering wall extend to s~gnaficant depths In order to judge the Importance o f basalt weathermg on the geochemical budget o f the major elements, the annual contnbutaon per cm 2 of ocean floor of a 1

277

cm deep layer o f basalt is compared to annual dissolved stream supply m table 3 The basalt weathering values were derwed from the age curves m fig. 2 Column 3 gwes the depth o f basalt weathering necessary to equate the weathering contrabutlon to stream supply. The effects of sedament cover in preventmg the chemical exchange between sea water and ocean basalts must be further studied In the long term balance o f the oceans, sea water m a y only be an antermedmte reservoir for cations exchangang between basalts and sediments, the net end effect would be the same whether cation exchanged directly between sedtments and basalt or entered actave oceanic circulation F o r the present tame, since we know one third of the ocean floor is vartually free o f sediments, the figures for basalt weathering contrabutlon to sea water an table 3 can be cut by one third m order to get an adea o f the maxtrnum effect o f sedament cover Tltamum as the element most hkely to be controlled by the basalt weathering process, an fact, given the present estLmates of stream supply, it as difficult to provide a source for the amount of titanium apparently enrxched m basalts dunng the weathenng process The iron balance o f the oceans as certainly strongly influenced b y sea water action on basalts and a source of iron must be looked for. Manganese and phosphorus reqmre greater weathering depths than do either tatanlum or ~ron, but they too m a y be strongly governed by weathering of basalt Silica, magnesmm and potassmm all would apparently require still greater depths of basalt weathering and our present lack o f knowledge o f the ocean floor prohlbats us from making further elaborataons on these elements The calcium and sodium concentration o f sea water seem the least lukely to be affected by basalt weathering

4. Conclusions Deep ocean basalts act as sources for slhca, magnesmm and calcmm and act as smks for titanium, potassmm, phosphorus, manganese, total aron and sodmm relatwe to sea water. Aluminum shows no conclusive exchange with sea water Ferric aron replaces ferrous iron The chemical exchange between sea water and basalts should be consadered m petrologac theories for

R HART

278

Table 3 Comparison of basalt weathering and stream supply

Chemical constituent

Annual weathering contribution from 1 cm layer of basalt

Annual stream*

Weathenng

contnbutxon

depth m cm

S102

+74

+1 3 X l 0 s

T102

-1 6

+0 3 X 102

17 600 18

(Fe)

-3 3

+(0 6 X 103)

182

Mn Mg

- 0 06 +2 3

+0 7 X 102 + 0 4 X 105

1 166 17 400

Na

- 0 21

+6 3 X 104

300 000

Ca

+2 1

+1 5 X 10 s

71 500

K

-0 9

+2.3 X 10 4

25 500

P

- 0 14

+0 2 X 103

1 430

Plus (+) indicates addition to sea water, minus ( - ) indicates extraction from sea water Units 10-9 g/cm 2 of sea floor Fourth column mdlcates depth of weathermg requtred to equate stream supply and weathering contribution * From Tureklan [39] t h e origin o f o c e a m c rocks a n d m t h e o r i e s c o n c e r n i n g the g e o c h e m i c a l mass b a l a n c e o f the o c e a n s

Acknowledgements Bert T a n n e r a n d T r o y H o l c o m b e v e r y h e l p f u l l y w r o t e the c o m p u t e r p r o g r a m s . E l l y n J o h n s o n a n d Gloria Shull w o r k e d o n the m a n u s c r i p t P e t e r V o g t , L e o n a r d J o h n s o n , S t a n l e y Hart, Eric S c h n e i d e r a n d Bill M o o r e all o f f e r e d h e l p f u l discussions d u r i n g t h e course o f t h e w o r k a n d m a n y useful criticisms o f t h e manuscript

References [1] A R.McBtrney and I G Gass, Relations of Oceamc Volcanic Rocks to Mid-Oceanic Rises and Heat Flow, Earth Planet Scl Letters 2 (1967) 265 [2] A E J Engel, C G Engel and R G Havens, Chemical Characteristics of Oceamc Basalts and the Upper Mantle, Bull Geol Soc A m , 76 (1965) 719 [3] H S Yoder and C E.Tdley, Ongm of Basalt Magmas, An Expernnental Study of Natural Synthetic Rock Systems, J Petrology 3 (1962) 342 [4] F Aumento, Magmatlc Evolution on the Mid-Atlantic Ridge, Earth Planet Scl Letters 2 (1967) [5] D H Green and A E Rmgwood, Mineral Assemblages m a Model Mantle Composition, J Geophys Res 68 (1963) 937

[6] R Kay, N J Hubbard and P W Gast, Chemical Characteristics and Ongm of Oceamc Ridge Volcamc Rocks, J Geophys. Res 73 (1970) 5925 [7] H H Hess, History of Ocean Basins, Geol Soc A m , Buddlngton Volume (1962) 599 [8] R S Dletz,Contment and Ocean Basra Evolution by Spreading of the Sea Floor, Nature 190 (1961) 854 [9] F J Vine and D H Matthews, Magnetic Anomalies over Oceamc Ridges, Nature 199 (1963) 947 [10] C G Engel, R L Fisher and A E J Emgel, Igneous Rocks of the Indian Ocean Floor, Science 150 (1965) 605 [ 11 ] G D Nlcholls, Environmental Studies m Sedtmentary Geochemistry, Scl Prog 51 (1963) 12 [ 12] J G Moore, Rate of Palagomtlzatlon of Submarine Basalt Adjacent to Hawan, U S Geol Survey Prof Paper 550-D (1966) D163-D171 [13] S R Hart and A J Nawalk, K, Rb, Cs, and Sr Relationships m Submarine Basalts from the Puerto Rico Trench) Geochtm Cosmochma Acta, in press [14] S H Hart, K Rb, Cs Contents and K/Rb, K/Cs Ratios of Fresh and Altered Submarine Basalts, Earth Planet Scl Letters 6 (1969) 295 [ 15 ] A F Mlyashlro, F Shldo and M.Ewmg, Dwerslty anc) Ongm of Abyssal Theolnte from the Mid-Atlantic Ridge Near 24 ° and 30 v North Latitude, Cont Mineral. and Petrol 23 (1969) 38 [ 16] T P Paster, Petrologic Variations Within Submarine Basalt Pillows of the South Pacific-Antarctic Ocean, Ph D Thesis, Florida State University, 1968 [ 17] A J Nawalk, Geology of the North Wail of the Puerto Rico Trench, Pit. D Thesis, University of Plttsburg, 1967 [18] A E J Engel and C G Engel, Basalts Dredged from the North-Eastern Pacific Ocean, Science 140 (1963) 1321

CHEMICAL EXCHANGE BETWEEN SEA WATER AND DEEP OCEAN BASALTS [ 19] I D Mutt and C.E.TtUey, Basalts from the Northern Part of the Rift Zone of the Mid-Atlantic Ridge, J. Petrol 5 (1964) 409 [20] F Aumento, The Mid-Atlantic Ridge near 45 ° North II From the Area of Confederation Peak, Canad J Earth Scl 5 (1968) 1 [ 21] C G.Engel and R.L.Flsher, Lherzohte, Anorthoslte, Gabbro and Basalt Dredged from the Mid-Indian Oceanic Ridge, Science 166 (1969) 1130 [22] E Bonattl, MechanlmaS of Deep-Sea Volcamsm m the South Pacific, Researches m Geochemistry, 2 J.Wfiey and Sons, New York, N.Y., 663, In Abdson 453 [23] H.G.Melson, G Thompson and T H.van Andel, Volcanism and Metamorphism m the Mid-Atlantic Ridge, 22 ° North Latitude, J Geophys. Res. 73 (1968) 5925 [24] A E J Engel and C.G Engel, Igneous Rocks of the East Pacific Rise, Science 146 (1964) 477 [25] F D.Mutr and C E Ttlley, Basalts from the Northern Part of the Mid-Atlantic Ridge J. Petrol 7 (1964) 193. [26] G D.Nlchols, A J Nawalk and E.E Hays, The Nature and Composition of Rock Sample Dredged from the MldAtlantac Ridge Between 22 N and 52 ° N, Marine Geol 1 (1964) 333. [27] A E.S.Engel and C G Engel, Composition of Basalts from the Mid-Atlantic Ridge, Science 144 (1964) 1330. [28] J.R~2ann and F.J Vme, An Area on the Crest of the Carlberg Ridge, Petrology Magnetic Survey, Phil Trans. Roy Soc. A 258(1965) 198 [29] F Aumento and B.D.Loncarevlc, The Mid-Atlantic Ridge Near 45 ° N III Bald Mountain, Canad. J Earth SCl. 6 1 (1962) 11.

279

[30] A Poldervaart and J.Green, Chemical Analysis of Submarine Basalts, Am. Miner 50 (1965) 1723 [31] C G Engel and T.E.Chase, Composition of Basalts Dredged from Seamounts off the West Coast of Central America, V S G S Prof Paper 525-C, C161-C163, 1965 [32] P Vogt, E D Schneider, O.Anderson and O Avery, Discontinuities m Sea Floor Spreading, Tectonophyslcs, m press. [33] J.P Hetrtzler, G O Dickson, E M.Herron, W C Plttman, III, and X LePlchon, Marine Magnetic Anomahes, Geomagnetic Field Reversals, and Mottons of the Ocean Floor and Continents, J. Geophys Res 73 (1968) 2119 [34] P.E Hayes and W C Pltmarm, III, Magnetic Lmeatlons m the North Pacific, North Pacific Vol. GSA, m press [35] A.Cox, R.R.Doell, G.B Palrymple, Tune Scale for Geomagnetic Reversal, m The History of the Earth's Crest, R.A Rmney, ed. (Prmceton Unwerslty Press, Prmceton, 1968) [36] A F Maxwell, Recent Deep Sea Drilling Results from the South Atlantic, AGU abstract, April, 1969 [37] G A MacDonald and T Katsuma, Chemical Composition of Hawauan Lavas, J Petrology 5 (1964) 82 [38] F.C Loughman, Chemical Weathering of the Silicates (Elsevier Pubhshmg C o , New York, 1969) [39] K K Turekmm, The Oceans, Streams and Atmosphere, m Handbook of Geochemistry, Wedepohl, ed (SprmgerVerlag Publ. C o , New York, 1969) [40] V Manson,Geochemlstry of Basaltac Rocks Major Elements, in Basalts The Poldervaart Treatise on Rocks of Basaltic Composition, Vol 1, H H Hess and A Poldervaart, eds. (Intersclence, New York, 1967)