basic lavas and shallow intrusionsmws vws vws wws ws ss ws 16.0 a 145 7.8-7.o 5.0 4.52 426 3.21 259...

THE AMERICAN MINERALOGIST, VOL 43, JANUARY_FEBRUARY 1958

ALTERATION OF OLIVINE AND ORTHOPYROXENE INBASIC LAVAS AND SHALLOW INTRUSIONS

H. G. Wrrsurur, Uniaersily of Sydney, Syd.ney, Australia.

ABSTRACT

Alteration products of olivine and orthopyroxene, previously identified as iddingsite,

bowlingite, and serpentine among other minerals, have been found to consist predominantly

of mixed layer smectite-chlorite with a variety of accessory minerals. Alteration products

from shallow basic intrusions may contain antigorite or chrysotile as well as smectite-

chlorite, and are much richer in MgO than those from basic lavas. Optically isotropic

varieties (chlorophaeite) are closely associated with smectite-chlorite minerals and some-

times have a poorly developed chlorite (?) structure.It is probable that all of these secondary minerals can form by deuteric alteration, but

evidence is presented which suggests that oxidized varieties (commonly containing goethite)

may also form by weathering of green varieties. Implications of chemical and physical

properties are discussed and it is concluded that equal volume replacement of Mg-rich

olivine or orthopyroxene in basic lavas results in selective leaching of Mg which may be

removed from the altered rock and recombined in the same secondary minerals in joints.

Quantitative data for the efiects of alteration on bulk composition of the enclosing rock

are presented. In shallow intrusions, where there is little evidence of Fe-enrichment in

residual solutions, complete solution of olivine is possible. Application of these results to

some problems in soil clay mineralogy is briefly considered.

INrnooucrroN

Although many optical, chemical ar'd tc-ray data are now available for

alteration products of olivine and orthopyroxene in igneous rocks, littleattempt has been made to organize these data, with the consequence thatrelations among alteration products are not clearly understood and mis-

identifications based on inadequate data are common. The object of thisstudy is to identify typical alteration products by r-ray techniques, toattempt an integration of ideas concerning their mode of origin and rela-tionships, and to discuss possible implications of these relationships. A

sampling problem is recognized in view of the widespread occurrence ofthese minerals and has been accounted for to some extent by selectionof material from widelv seDarated localities.

Pnnvrous Wonx

It is impossible to be exhaustive in the treatment of previous work' for

the literature on these minerals is very large. Among the more importantcontributions are those of Ross and Shannon (1926) whose chemicalformula of iddingsite is quoted in most textbooks, Peacock and Fuller(1928) who have presented the most comprehensive study of the occur-

rence and mode of origin of chlorophaeite, and Caillere (1935) and Cail-

128

ALTERATION OF OLIVINE AND ORTHOPVROXENE

lere and Henin (1951) whose r-ray and thermal studies of bowlingite

have led to redesignation of this mineral as saponite.

nlBrnoos on INvnsrrcarroN

X-ray powder photography using a Philips X-ray Diffraction unit was

the principal method used in identification. Filtered copper radiation was

found adequate except with optically isotropic material. Some basal spac-

ing expansions were obtained from oriented aggregates using a Philips

Geiger Counter Spectrometer with filtered copper radiation.Refractive index measurements were made with standard immersion

oils which were checked with a Leitz-Jelley refractometer. Olivine compo-

sitions were determined from n, using Kennedy's curves (Troeger, 1951,

p.37).Refractive indices of the alteration products are quoted only to the

nearest second decimal place because the index varies with the type of

immersion oil used.Plagioclase compositions were determined with the universal stage

microscope using Turner's (1947) method and van der Kaaden's (1951)

data. All measurements of 2V were also made with the universal stage

microscope. All other mineral identifications and fabric descriptions were

made with a standard petrographic microscope.

MolB ol OccunnBNCE AND IlnNru'rcarrort

In basic lavas olivine and orthopyroxene are commonly altered to a

reddish-brown substance with high birefringence and refractive indices

ranging from approximately 1.61 to 1.86. This is usually identif ied as

iddingsite if it is pseudomorphic after olivine, and is assigned the chemical

formula, MgO.FezOr'3SiOz.4H2O (Ross and Shannon, 1926). Average

chemical analyses of iddingsite are given in Table 4A,' Optically identical

material in amygdules and joints is sometimes called chlorite. A recent

*-ray investigation of iddingsite (Nling-Shan Sun, 1957) revealed the

presence of goethite mixed with amorphous substances.A green substance with moderate birefringence and refractive indices

ranging from approximately 1.48-1.62 is also a common alteration prod-

uct of olivine and orthopyroxene and is exceptionally abundant in shallow

basic intrusions. Optically identical material occurs in amygdules andjoints. This material has been assigned a wide variety of mineral names

including bowlingite, serpentine, chlorite, nontronite, saponite, and

vermiculite. Average chemical analyses, all of saponite from shallow in-

trusions, are given in Table 4A. X-ray and thermal studies by Caillere

and Henin (1951) indicated a montmorillonite-type structure and this,

combined with the chemical composition, Ied them to redefi.ne bowlingite

as saponite. Other r-ray powder studies of minerals fitting the optical

121

122 H, G. WILSEIRE

Tasln 1. x-nav Dera: Seuprrs UNrnrernl

Group 1

d

l s 4 9 44 5 24 . 3 03 . 3 73 . 2 62 59-2 . 5 21 7 11 . 5 1 3

rdtI *

K

K

m w Ks Kvs G; Q

vs G; Q;v w Kv w K

v w Gm Km s G ; Qv w G ; Kvw G; Q;vw G; Kn v Q ; Kw Q ; Kw G ; Kw Q ; KM W Q ; Km w G ; Kw G ; Kw Q

7 3 A7 . O4 . 4 74 . 1 93 . 3 53 . 2 22 . 7 22 . 6 62 . 5 52 . 4 32 2 82 2 42 . 1 92 1 21 . 8 21 7 11 6 71 544I .4951 4591 3 7 7

m s Ss SW F ?w Q ?w F ?

m Sw Sm w S

1 4 . 0 4 m S4 . 9 2 w G ; S4.49 ms S4 l) vs u. t /1 lm r

3 . 3 3 w Q3 . 2 O m F2 . 7 O m G ; S2 . 5 9 m S2 . 5 0 w S2 . 4 4 s G2 . 2 5 w S2-15 vw S1 . 9 6 v w S1 . 7 9 w S1 7 2 s G ; S1 . 6 0 w S1 . 5 6 3 w S1 . 5 1 0 m w S1 . 4 6 0 w S

1 4 . 0 4 m S4 . 8 0 w S4 . 4 6 m S4 . 1 8 s G3 7 0 w F3 3 5 v w a3 2 1 m F2 9 4 v w F2 . 6 9 m G2 . 5 9 w S2 5 2 w S2 4 4 m s G2 . 2 2 w S1 . 7 1 m G1 . 5 6 v w S1 . 5 0 9 w S1 4 5 5 w S

' I :intensity, visually estimated.t Id:_identif ietion. S:smectite-chlorite; F:feldspar; Q:goethite; Mg:magnetite; Ca:calcite; Q:

quartz; Mi :mica; T : talc; Se : sepiolite; Sp :sslp.rtite; K :kiaolinite.1. Carmel Bay, California. Pseudomoi.phs alte; olivine.2. Carmel Bay, California. Pseudomorphs after olivine.3. Santa Rosa, California. Pseudomorphs after olivine.4. Orange, New South Wales. Joint-fl i ing.

Id

l s . 0 41 4 . 0 s S7 . 4 -7 O s S ; S p4 . 8 3 m S4 . 5 7 m S3.84 w Ca3 . 6 1 m S ; S p3.03 vs Ca2 . 8 9 w S2 . 6 5 -2 . 4 1 w S2.5O mw Ca; S ; Sp2 2 8 m C a ; S2 . O 9 m C a1 . 9 1 m s C a1 . 8 7 m s C a1 . 7 4 v w S1 . 6 3 w C a1 . 6 0 w C a1 . 5 4 0 s S1 . 5 0 6 m w S1 4 4 3 w C a1 4 1 8 w C a1 3 2 6 w S1 2 9 9 w S1 . 2 3 0 w S

v s S

v w Sw Ss S

m Fs Sm w Sv w Sv w Sw w Sw S

s Sw S

1 6 . 0 A1 4 57 . 8 -7 . O5 . 04 . 5 24 2 63 . 2 12 5 92 . 4 72 . 2 82 2 01 .901 . 7 21 . 5 3 31 5121 . 3 1 4

1 4 . 0 A v s S7 - 1 3 v w S4 . 6 0 m S4 2 5 w a4 . O 2 w F3 3 3 w Q3 1 7 s F3.03 w Ca2 9 5 w F2 8 8 m F2 . 8 3 w S2 . 7 1 w S2 . 5 4 w S2 . 5 O w S2 . 4 7 w S2 . 1 2 w S2 . O 9 w S2 . 0 5 v w S1 . 9 9 w S1 9 6 w S1 7 4 v w S1 . 5 4 6 w w S1 . 5 3 1 m S1-487 w S1 . 4 7 2 w S1 ..387 vw S

Group 2

5. Ebbetts Pass, Cali lornia. Pseudomorphs alter hvoersthene.6. Orange, New South Wales. Joint-6ll ing7. Carmel Bay, Cali lornia. Pseudomorphi after olivine.8. Prospect Hil l, Sydney, New South Wales. Joint-fit l ine.

ALTERATION OF OLIVINE AND ORTHOPYROXENE

T.q.srn 1. (conti.nued)

123

Group 3

IdId

v s Ss S ; S pv w Sm w Sw Qm S ; S pm Qw C av w Svw S; Sps Sv w S

1 5 0 -1 4 . 07 . 3 04 . 8 74 . 5 84 3 23 . 6 13 . 3 33 . 0 32 . 6 62 . 4 81 . 5 4 51 . 5 1 4

SMi

s ; SpMiSs . q h

MiIF

SSs; SpSSSSSSS

SSS

1 5 0 A1 0 . 07 . 4 -7 . O5 . 04 . 5 53 . 6 03 . 3 s3 1 92 . 9 92 . 9 02 . 6 22 . 5 12 . 4 52 2 02 1 32 . O 51 8 31 7 61 6 21 540-1 . 5 2 01 . + 2 51 . 3 1 5

r 4 . 0 A w S9 . 3 5 w T7 5 -7 .O vs S; Sp4 . 5 7 s S3 8 9 w3 . 6 2 s S ; S p3 1 3 w T3 . 0 3 w C a2 . 8 5 w S2 . 6 6 -2 . 4 1 w S2 5 0 v s S ; S p2 3 3 v w S2 2 8 v w S ; C a2 . 1 5 m S2.O9 vw Ca; S1 9 6 w S1 . 9 1 v w C a ; S1 8 7 w C a ; S1-79 mw S1 . 7 4 w S1 . 6 4 v w S1 . 5 6 w S1 . 5 7 3 s S1 . 5 0 3 s Sl 4 l 2 w S1 3 2 9 w S1 ..307 m S1 . 2 7 8 w S1 2 5 2 v w S1 . 1 6 7 w S

1 4 . 5 A v s SI 2 . l O w S e ?9 . 5 0 w T7 . 5 -7 . O s S ; S p5.88 w Se?4 . 9 O m S4 . 5 7 s S ; S e ?3 . 6 6 s S ; S p3 - 2 5 m T2 . 9 O m s S2 6 5 w S2 . 5 8 2 S ; S p2 4 3 s S2 - 2 7 m S2 0 8 -2 O O w S1 . 8 8 v w S1 7 4 w S1 7 1 v w S1 5 7 6 w S1 . 5 4 0 s S1 .508 mw S1 . 4 0 1 w S1 3 2 4 w S1 2 9 9 v w S

s

s

S

m

mw

m

S

S

9. Prospect Hill, Sydney, New South Wales. Joint-lilling.10. Prospect Hil l, Sydney, New South Wales. Joint-fi l l ing.11. Prospect Hill, Sydney, New South Wales. Pseudomorphs after olivine.I l. Dumbuck Quarry, Milton, Scotland. Joint-6ll ing.

Group 4

Id

1 5 0 - A1 4 . 0 s4 . 5 v w4 4 8 w2 5 9 v w1 . 5 5 v w

sS

S

s Ss S ; S p

W Smvs S; Sp

w C aw SW Svw S; Spvw S; Cam Sv w Sv w Sv w Sw S

1 4 . 5 -14 .O7 . 15 . 3 74 5 74 . 2 63 . 5 83 . 1 43 0 5

2 . 6 52 5 12 . 2 51 .83I . 6 31 5 8I . 5 3 91 .430

1 4 . 5 A v s S7 . 2 s S p ; S4.81 vw S4 . 5 9 w S4 . 2 6 v w O3.62 m Sp; S3 3 3 m Ot . 7 i -1 7 1 w S1 . 5 4 1 s S

1 5 0 -1 4 04 . 4 52 . 6 01 . 5 3 5

Group 3

l.t. Bowling Quarry, Dumbartonshire, Scotland. Joint-fi l l ing.14. Bowling Quarry, Dumbartonshire, Scotland. Joint-fi l l ing.15. Localitv unknown. Vesicle-fi l l ine.16. Walla Walla, Washington, Vesiciefilling.

124 H. G. WILSHIRE

description or mode of occurrence of this material are those of Bradley(1945), (vermiculite), Prider and Cole (1942), (nontronite), and Earley

and l\ l l i lne (1956), (regularly interstratif i .ed montmoril lonite-chlorite)'

When the alteration product, whether it occurs as pseudomorphs after

mafic minerals, joint-filling or in amygdules, is optically isotropic it is

usually called chlorophaeite (Peacock and Fuller, 1928). It is charac-

terized by a refractive index ranging from 1.50 to I.62, optical isotropy,

and a bright orange to deep green color. Average chemical analyses are

given in Table 4A.

INtBnpnnrerroN oF rrrn X-R'c,v PowoBn Puorocnapus

The data presented in Tables l, 2, and 3 are representative of 170

powder photographs of 60 specimens. Sample numbers are arranged so

that the same number refers to material from the same bulk sample.

These are divided into four groups, nos. 1-4, which represent red, oxi-

dized alteration products usually called iddingsitel nos. 5-7, which repre-

sent green alteration products in lavas; nos.8-14, which represent green

alteration products in shallow intrusions; and nos. 15-16' which repre-

sent optically isotropic alteration products.

Group I

Photographs of untreated material (Table 1) are characterized by a

broad but strong reflection near 14 A. Rarely very weak second and third

order reflections from the basal spacing occur. A few specimens have a

reflection near 7 A with no 14 A reflection. A rveak 060 reflection typi-

cally occurs between 1.49-1.51 A, although two lines are frequently pres-

ent in the region (1.49-1.544; in which the 060 reflection for dioctahedral

Tesln 2. X-n.lv Dlre: Basar, SpecrNcs ol Gr,vcor,arrn Saupr,rs

Sample Number

n1mws

sVS

VS

VS

sS

lnS

I

J

6789

101 1l 2I J

t4

1s .33 .{1 5 . 81 6 . 3 51 6 . 0 51 6 . 515.22r . ) . 11 4 . 51 5 . 51 5 . 51 5 . 575.77

ALTERATION OF OLIVINE AND ORTHOPYROXENE

and trioctahedral structures occurs. With e_thylene glycol treatment(Table 2) the basal spacing expanded to 15-16 A. Upon heating to 500o C.,

a new line appeared between 9.5-10 A in specimens having a 14 A spac-

ing (untreated). Heating to 700" C. (Table 3) produced no additional re-

sults. Higher order reflections from the basal spacing and the 7 A struc-

ture disappear with heating to thes^e temperatures.The usual expansion of the 14 A spacing with glycol solvation and

collapse with heating indicate the presence of vermiculite or smectite.x Apossible means of distinguishing these by Mg-saturation and glycerol ex-

pansion was suggested by G. F. Walker (Pers. comm.). No change in ex-

pansion was observed after a 24 hour treatment with saturated l'IgClz

solution indicating that the structure is a smectite. Since the expansion

is rarely to the full l imit (17 A) expected of glycol solvation, it is possible

that the smectite is randomly interlayered with chlorite. Absence of

higher order reflections, which render this interpretation difficult to eval-

uate, may be due to auto-oxidation. As will be shown later, alteration

takes place before consolidation of the magma is complete so that high

temperatures and, in lava flows, oxidizing conditions may be expected.

It is concluded, therefore, that oxidized alteration products are composedpredominantly of smectite-chlorite. Goethite is almost always present,

qtartz and calcite are common constituents and talc and mica rare con-

stituents.

Group 2

These powder photographs (Tables 1-3) are nearly identical with those

of group 1 and are interpreted in the same way. The more common oc-

currence of higher order reflections from the basal spacing is attributed

to incomplete oxidation of iron (which also accounts for optical differ-

ences). Evidence presented in a later section suggests that weathering of

green alteration products can result in extraction and oxidation of iron,

producing aggregates which are optically identical with oxidized varieties,

this process possibly accounting for differences in position of the 060 re-

flection between the two. Calcite is common in these aggregates, while

magnetite, quartz and feldspar sometimes occur. Talc is a rare constitu-

ent .

Group 3

Powder patterns of untreated material (Table 1) from shallow intru-

sions are characterized by a strong reflection between 14 A and 15 A with

nonintegral higher order reflections from this spacing of comparable in-

x The term smectite (Brown, 1955) denotes members of the montmorillonite group.

125

126 H. G. WILSHIRE

tensity. The 060 reflection is usually strong and near 1.544, although itis usually accompanied by a weaker l ine near 1.50 A. With ethyleneglycol treatment (Table 2), the 14 A spacing expanded to 15-16 A andgave rise to a nonintegral series of higher order lines. Upon heating to500' C-. a new line appeared between 9.4 10 A in most specimens, at11.78 A in one specimen. Heating also resulted in a considerable reduc-tion in intensity of higher order reflections from the basal spacing so thatthese patterns are, with the exception of differences in intensity and posi-tion of the 060 reflection, identical with most heated specimens of Group1. Since heating was done in air, oxidation of iron produces from the green

Tlsr,r 3. X-nav Dnu: Be.sn SplcrNcs ol Hrarnn Snlrprrs

23 hr 700" c. 23 hr . 700 'C. 2 1 h r . 5 2 5 " C . 23 hr. 700'C 25 hr. 700'C 2hr .710" C.

1 z 3 6 7 8

I d I d I d T d I d l

m 9.77 mw 9 . 5 v s 9 . 6 0 v s 9 . 9 3 s 9 . 5 v v w

700" c 23 hr . 525 'C. 23 hr . 700 'C 2 h r . 7 7 O ' C t hr. 480' C

9 10 t2 l . l 1 4

T d I d d I d T

VS 9 . 4 5 m 9.6 s 9 .40 vvw 1 0 . 4 w

9 . 8 2

23 hr

7 1 . 7 8

alteration products an aggregate which is optically identical with naturaloxidized types. Heating to 700o C. (Table 3) produced no additional re-sults.

This behavior again indicates that an expanding 14 A structure-is pres-ent. That this is probably interlayered with a non-expanding 14 A struc-ture is indicated by the nonintegral series of higher order reflections. Itis also possible that variable amounts of smectite-chlorite are mechani-cally mixed with a kaolin-type structure (antigorite or chrysotile), theIatter giving rise to strong reflections near 7.2 A and 3.6 A. In specimensshowing large expansion with glycol solvation, the second order reflec-tion split, one line being roughly half the spacing of the expanded spac-ing, the other occurring at about 7.2 A, which strongly suggests a mechan-ical mixture of a serpentine mineral with the smectite-chlorite. An at-tempt was made to separate the two, but without success. Although

ALTERATION OF OLIVINE AND ORTHOPYROXENE 127

Ifargreaves and Taylor (1946) found that conversion of chrysotile andantigorite to olivine takes place at about 600o C., no reflections belongingto olivine were found in photographs of material heated to and above thistemperature.

Material from shallow intrusions is therefore composed at least in partof smectite-chlorite with possibly one or both of the serpentine minerals.Considerable variation in proportions of these components occurs amongdifferent samples. \'Iinor amounts of talc, mica, quartz and possiblysepiolite are sometimes present.

Group 4

Optically isotropic alteration products are closely associated withanisotropic material which is optically identical with members of Group1. In addition, optically isotropic material is itself sometimes submicro-scopically crystalline, giving diffuse reflections corresponding to thestrongest l ines in powder photographs of Group 1. No change in positionof the basal reflection was observed with heat and glycol treatment, sothat it is probable that the incipient structure is chlorite. Absence ofhigher order reflections may again be due to auto-oxidation of iron.Goethite and calcite (optically detectable) are sometimes present.

DrprBnnNrrar, Tunnuel Axarysrs

With the exception of specimen no. 14, thermograms (Fig. 1) of speci-mens from shallow intrusions all have endothermic peaks between 100-200" C. which are attributed to loss of adsorbed water from the smectitecomponent. Absence of this peak in no. 14, which also has a smectite com-ponent, is not explained. The second endothermic peak, representing lossof structural water, is typically broad and occurs anywhere between535'-700' C. It is likely that this behavior reflects variation in the pro-portions of chlorite and smectite. In most specimens this peak is fol-lowed by another, relatively sharp endothermic peak, representing de-composition of the aggregate, between 780-840" C. In specimens 13 and14, this peak is followed by another endothermic peak at 865o C. and820" C., respectively. Doubling of the last endothermic peak probablyindicates decomposition of different components of the aggregate at dif-ferent temperatures. Most samples show a final exothermic peak between820"-860o C. indicating the formation of new mineral phases.

The thermogram of no. 6 (from a lava) is very similar to that oI non-tronite. No. 4 is the weathered equivalent of no. 6. ' Ihe endothermic peakat 305o C. in the latter is due to decomposition of goethite, while the peakat 545o C. may be due to kaolinite.

r28 H, G,'WILSHIRE

tooDEGREES C

Frc. 1. Difierential thermal analyses.

tooo

ALTERATION OF OLIVINE AND ORTHOPYROXENE 129

Pnrnor-ocv

Specimens of basalt flows containing both red and green varieties of

alteration products were obtained from Carmel Bay, California (type

Iocality for iddingsitel Lawson, 1893), from near Santa Rosa' California,

Hawaii, Orange, New South Wales, and from Tertiary lavas near Eb-

betts Pass, California.

Carmel Bay, CaliJornia

At Sunium Point a series of basalt flows outcrop with an exposed thick-ness exceeding 75 feet. Weathering to a depth of approximately 20 feet

has resulted in bleaching of the groundmass and alteration of pyroxene

and plagioclase phenocrysts. The only obvious mafic mineral is redpseudomorphs after olivine. In very deeply weathered material pseudo-

morphs after olivine may be detected only with a hand lens. Below the

weathered zone the groundmass is black, fresh pyroxene phenocrysts may

be seen with a hand lens, and, at the base of the exposure, the alterationproduct of olivine is green.

Weathered rocks are characterized microscopically by an abundance

of smectite-chlorite and goethite pseudomorphs after olivine, plagioclase

and pyroxene phenocrysts completely replaced by a colorless to pale

green fibrous material with moderate birefringence, set in an intergranu-lar groundmass of altered plagioclase microlites, granular and euhedralpyroxenes, iron oxides and unaltered apatite. Small amounts of pale

green to brown interstitial material is often present, while biotite occursrarely.

Almost all of the specimens show a distinct flow banding in the plagio-

clase microlites. There is commonly a concentration of very finely crys-talline green to brown material and iron oxides in bands parallel to theflow banding, appearing as a distinct macroscopic lineation on cut sur-faces. Pseudomorphs after olivine within these bands are distorted (Figs.

2 and 3) .In unweathered specimens the same textural relations hold, but olivine

is usually the only altered mineral. Augite (2V 45'-48") and plagioclase(An6-ur) aggregates are common. In some specimens the groundmasspyroxenes consist in part of colorless euhedra of orthopyroxene which arepartly altered to green smectite-chlorite (l). Iron oxides occasionally pre-

dominate over clay minerals in pseudomorphs after olivine, while in sev-eral specimens green and red alteration products occur side by side in thesame pseudomorphic aggregate (Fig. a). Pseudomorphs are usually con-nected by delicate stringers, which appear to be thin films coating plagio-

clase and pyroxene microlites, of material optically identical with that in

the pseudomorphs, and sometimes have halos of the same secondary min-

erals.

130 H. G, WILSHIRE

Frc. 2. Pseudomorph after olivine distorted by flow of the enclosinglava. Plane polarized light. X45.

Rocks from another locality about a mile east of Sunium Point shownearly identical textural and mineralogical relations, except that augite(2V 46"-54')-plagioclase (Aruo_ur) aggregates are usually Iarger thanthose in the Sunium Point flows, and, although often very deeply weath-ered, colorless clay pseudomorphs after these aggregates are obvious in

Frc. 3. Smectite-chlorite, iron oxide,florv. Color banding is opposite

biotite pseudomorph after olivine distortedfrom usual. Plane polarized light. X30,

by

,ILTERATION OF OLIVINE AND ORTHOPYROXENE

Frc. 4. oxidized ;*H:HT:ijffi ,i*

pseudomorphs af ter

hand specimen. Pseudomorphs after olivine are dull brown to nearly

colorless and, when nearly colorless, are distinguished from weathered

pyroxene and plagioclase only by their form.

In thin section, the material replacing pyroxene-plagioclase aggregates

is colorless to pale green, has a moderate birefringence and occurs as

randomly oriented fibers in the pseudomorphs.

S attta Rosa, C alifornia

All of these specimens are weathered and in some all the major minerals

are altered and pseudomorphs after olivine are bleached. \Iost specimens

are characterized by olivine phenocrysts (Foru-ru) partly altered to iron

oxides and smectite-chlorite, plagioclase (Anuo-un) and some augite

(2V 53"-57.) phenocrysts set in an intergranular groundmass composed

of plagioclase, iron oxide and clinopyroxene microlites with accessory

apatite. A small amount of optically isotropic pale green or brown meso-

stasis is usually present. FIow banding in plagioclase microlites is con-

spicuous and parallel to this are bands in which the clay minerals and

iron oxides are concentrated. The bands are often connected by stringers

cf the same minerals resulting in a crudely rectangular or polygonal pat-

tern. Pseudomorphs after olivine within these bands are distorted and

alteration of olivine is complete. Within the polygonal areas olivine is in-

completely altered to green smectite-chlorite. Identical secondary min-

erals, including optically isotropic varieties, also occur in microscopic

amygdules in the bands (Fig.5). Weathering has produced the same

131

H. G. WILSHIRE

effect as in the Carmel Bay specimens. Cores of altered olivines are oc-cupied by a colorless, fibrous material which is randomly oriented.

Hawaiian basalts

Specimens, from the museum collection, University of California, con-sist of olivine phenocrysts (Fozr-eo) partly or completely altered to oxi-dized smectite-chlorite, iron oxide aggregates, augite (2V 52") or titan-augite and plagioclase (An26_ae) phenocrysts in an intergranular ground-mass composed of plagioclase, granular clinopyroxene and iron oxidemicrolites with accessory apatite. olivine microlites are present in somespecimens.

Orange, New South Wales

These specimens were obtained from a small quarry in a single basaltflow, the surface of which has been severely weathered. Part of the flowhas well developed columnar jointing and adjacent to each joint plane thebasalt has a dark rim approximately 0.5" wide in which olivine has beenaltered to smectite-chlorite and iron oxides. The same minerals also occuras thin fi lms coating the joint surfaces. For a distance of about 10 feet be-low the surface both the joint-filling and the altered rims are brightorange red.

The unaltered rock consists in part of augite (2V 48.-50.) phenocrystsand aggregates which ophitically or subophitically enclose plagioclaselaths (An5s-55) and often enclose rounded grains of olivine. In areas be-tween the evenly distributed pyroxene-plagioclase intergrowths, the rock

Frc. 5. Smectite-chlorite vesicle-filling. Plane polarized light. X45.

ALTERATION OF OLIV]NE AND ORTHOPYROXENE

consists of plagioclase Iaths and olivine phenocrysts (Foru-nu), some of

which are highly resorbed, set in a glassy groundmass. Some iron oxides,

rounded olivine grains, rod-shaped clinopyroxenes and skeletal plagio-

clase crystals are distributed through the glass. In the altered rims below

the weathered zone, olivine has been partly or wholly converted to green

smectite-chlorite; in a few cases (Fig. 6) red cores are surrounded by

green rims, but the opposite relations may be seen in the same thin sec-

Frc. 6. Pseudomorphs after olivine consisting of oxidized smectite-chlorite cores and green

smectite-chlorite rims. Plane polarized light. X115.

tion. Green smectite-chlorite also occurs as thin films along cleavages andfractures in plagioclase and clinopyroxene grains and along grain bound-

aries (Fig. 7). Such veinlets disappear at crystal-glass boundaries and theglass is locally altered, the veinlet reappearing in crystals on the opposite

side of the altered glass. In a few places these veinlets occupy micro-shear

zones (Fig. 8). Within the weathered zone joint-fi l l ing and alterationproducts of olivine and glass are red aggregates of kaolinite and goethite.

With a few mineralogical and textural difierences, this occurrence is

nearly identical with that recorded by Smedes and Lang (1955).

Ebbetts Pass, C aliJornia

Green Smectite-chlorite is very widespread in the Tertiary lavas of theN{ehrten Formation as an alteration product of olivine in basalts and ofhypersthene in andesites. It is also present in amygdules and as coatings

on joint surfaces. In many places it occurs in isolated patches several

inches to several feet across within flows, and in these occurrences its dis-

133

134 H. G. WILSHIRE

Frc. 7. Smectite-chlorite veinlet. Note control of grain boundaries and fractures on position

of the veinlet Plane polarized light. X105.

tr ibution shows no relation to weathered surfaces. Coats (1940) men-tioned some of these occurrences and referred to the alteration product asbastite. In view of the common retention of hypersthene cores, there is nodoubt that this mineral alters to a product optically identical with thatwhich replaces olivine. The andesitic lavas also commonly have ortho-

Frc. 8. Smectite-chlorite veinlets occupying micro-shear zone. Glass is locally alteredsmectite-chlorite. Plane polarized lisht. X 55.

ALTERATION OF OLIVINE AND ORTHOPYROXENE

Frc. 9. Pseudomorph after olivine and material extracted duringalteration. Plane polarized light. X345.

pyroxene microlites partly or completely altered to the same mineral,while the clinopyroxene microlites have remained unaltered. Aggregatespseudomorphic after hypersthene commonly contain calcite and some-times contain cristobalite.

Specimens of altered shallow basalt intrusions were obtained fromBowling Quarry, Dumbartonshire, Scotland (type locality of "bowling-ite"; Hannay,1877); from Dumback Quarry, Milton near Bowling; andfrom Prospect Hil l near Sydney, New South Wales.

Bowling Quarry, S cotland,

Nlaterial from this locality was all taken from joints where it occurs asveins rr-"-1|" thick. According to Hannay (1877) it also occurs as vesicle-filling and as pseudomorphs after olivine in the dolerite. Other constitu-ents of the rock are unaltered.

Dumbuck Quarry, S cotland.

This material has the same mode of occurrence as that from Bowling

Quarry. In thin section green smectite-chlorite pseudomorphs olivinemicrophenocrysts and occurs as interstitial filling and as thin films coat-ing other minerals (Fig.9). It is commonly intergrown with a colorless,micaceous material. A few plagioclase phenocrysts (Anuo-ur) which en-close clinopyroxene grains occur along with the altered olivine in an inter-granular groundmass composed of plagioclase, clinopyroxene and ironoxide microlites with accessory apatite.

135

136 H. G, WILSHIRE

Prospect Hi.ll, liew South, Wales

Green smectite-chlorite is abundant in this dolerite intrusion, occur-ring as joint-filling, fibrous aggregates pseudomorphic after olivine andas interstitial filling. In the dolerite phases it is commonly associatedwith green and brown mica while in pegmatitic phases it occurs as inter-stitial filling and rarely as pseudomorphs after olivine. In the chilled mar-gin olivine phenocrysts (Foro-*) are fresh or partly altered, alteredphenocrysts being connected by veinlets of optically identical material.

Frc. 10. Resorbed plagioclase with sieve structure due to enclosedclinopyroxene and iron oxide. Plane polarized light. X95.

Weathered parts of the chilled margin show partial oxidation of altera-tion products, grading away from exposed surfaces into green alterationproducts. Such relations may be seen in a single thin section. Specimensused in this study were selected from joints and pseudomorphs afterolivine.

Specimens of basalt f lows containing optically isotropic alterationproducts (chlorophaeite) were obtained from the Grande Ronde dike,Oregon (Peacock and Fuller, 1928) and from several Tertiary lavas ineastern Washington and Oregon near Walla Walla, Washington.

Grande Rond.e d.ike

In view of some discrepancies between the author's observations andthose of Peacock and Fuller. soecimens from this localitv will be de-scribed in detail.

ALTERATION OF OLIVINE AND ORTHOPYROXENE

In the chilled margin a few olivine phenocrysts (Fos7) partly altered to

chlorophaeite and anisotropic smectite-chlorite, a few plagioclase aggre-

gates which are corroded and have a sieve structure enclosing clino-

pyroxene and iron oxide grains (Fig. 10), and rare qts'artz grains with clino-

pyroxene halos are set in an intergranular groundmass composed of

oli,rin" partly or completely altered to both red and green smectite-

chlorite and chlorophaeite, plagioclase, clinopyroxene and iron oxide

microlites with accessory apatite. There are a number of chlorophaeite

amygdules some of which contain carbonate. Amygdules in which the

chlorophaeite is bright orange are connected by a delicate network of

Frc. 1 1. Peripheral alteration of plagioclase resulting in formation of skeletal plagioclase

grains. Areas of high relief within the plagioclase are clinopyroxene. Plane polarized light.

x55.

veinlets which consist of chlorophaeite films coating the groundmass

minerals. These stringers form a crudely polygonal or rectangular pat-

tern. Amygdules which are not so connected contain green chlorophaeite.

The center of the dike consists of plagioclase phenocrysts and aggre-

gates (An12-aa) and olivine phenocrysts which are partly or completely

altered, set in an intersertal groundmass composed of skeletal plagioclase

grains, iron oxide microlites and glass or chlorophaeite with accessory

apatite. Plagioclase is commonly intergrown with clinopyroxene and is

peripherally altered, the alteration being controlled by cleavage planes

resulting in skeletal plagioclase grains (Fig. 11). Chlorophaeite amygdules

have a halo extending into the groundmass in which plagioclase has been

partly altered to chloroplhaeite (Fig. 12).

137

138 II, G. WILSHIRE

n l o o p o r O n r c r x

when the oxidized alteration product (iddingsite) was first named(Lawson, 1893) it was considered to be a primary mineral precipitateddirectly from basaltic magma. other investigations showed that residualolivine is commonly present within pseudomorphs and this hypothesiswas abandoned in favor of rddings' (1892) suggestion that it formed byweathering of iron-rich olivine. since the work of Ross and Shannon(1926) on iddingsite and that of peacock and Fuller (lg2$ on chloro-

Frc' 12. chlorophaeite amygdule with alteration halo. plane polarized light. x40.

phaeite, both of these have been generally accepted as products of deu-teric alteration of primary mafic minerals in basaltic lavas, iddingsitebeing a distinct mineral species, chlorophaeite being an amorphous sub-stance of variable composition. Green alteration products, whether theyare called bowlingite or any of their other names, are also rvidely thoughtto arise by the same process.

Ross and Shannon presented convincing arguments favoring a deu-teric origin which have been substantiated by subsequent investigations(e.g., Edwards, 1938; Fuller, 1938). However, Ming-Shan Sun (1g52)has suggested that identical alteration may be continued in weathering.This conclusion is based on the identity of deuteric alteration products(only goethite was identified) with weathering products and similarityof chemical changes involved in both processes. The occurrence of bothgreen and red varieties of alteration products at carmel Bay, california,Orange, New South Wales, and Prospect Hill, New South Wales whereoxidized material is present only in the weathered horizon crearly indi-

ALTERATION OT'OLIVINE AND ORTHOPYROXENE 139

cates that weathering may change the original deuteric alteration prod-

ucts rather than simply continue the same process. It was shown from

r-ray data that both alterat:ion products from Carmel Bay are composed

of the same structural types except that the oxidized variety containsgoethite. Conversion of grt:en smectite-chlorite (sometimes mixed with

magnetite) to red smectite-chlorite plus goethite can be effected either

by leaching iron and recombination as goethite or by weathering of in-

cluded magnetite, in either case the goethite acting as a pigment' Weath-

ering of the Orange, New South Wales material resulted in formation o{

goethite, but the smectite-,:hlorite did not survive and in its place kao-

linite is present. As wil l be shown later, the additional constituent,

smectite-chlorite, of "iddingsite" identif ied in this study may also form

by weathering of mafic mirrerals. However, the weathering products dif-

fer in being dioctahedral 'rhile oxidized smectite-chlorite in the same

rocks is usually intermedizrte.

Trlrn or AttBnerrorg

Edwards (1938) and Ross and Shannon (1926) presented textural evi-

dence which suggests that alteration takes place after the enclosing lava

ceased to flow, but sometimes prior to completion of crystall ization ofprimary minerals. Fresh olivine rims surrounding altered olivine cores

suggested to Edwards a se(fuence in which crystallization of olivine was

followed by hydration-oxitlation alteration, which in turn was followed

by renewed crystall ization of olivine. To Ross and Shannon, the same

relationships suggested zonal replacement of olivine which had com-

pleted crystallization before alteration commenced.Ross and Shannon furth,:r state that "The magma must have come to

rest before iddingsite formed for though it is a very brittle mineral it is

never fractured, or distortr:d by flow." Descriptions of the Carmel Bay'

California and Santa Rosa, California basalts showed that, in these rocks,

distortion of pseudomorphs is common, but restricted to narrow bands

parallel to the flow bandin5; (Figs. 2 and 3). These data are interpreted as

indicating at least partial alteration of olivine prior to cessation of flow,

followed by distortion in shear zones developed when laminar flow is

interrupted by consolidation.

Cupurcnr aNn Pnvsrc:er PnoprnuEs AND'fnBrn Ilrpttcarroxs

It is evident from the r-ray data that the formula of "iddingsite" re-

presents at least two minerals and possibly as many as three or four, and

that, in general, none of these are optically distinguishable.Notwithstanding mineralogical difficulties, bulk chemical composi-

tions are important, for they represent alteration products of primary

140 H. G. WILSHIRE

minerals lvhose composition can often be determined from relicts. It is,however, still a matter of conjecture whether or not relicts are repre-sentative of the entire original crystal. Most of the chemical data nowavailable are of "iddingsite," chlorophaeite, and associated olivine. Bycomparison of these analyses it is seen that, although a broad range ofcompositions occurs, they are characterized by high Fe/Mg ratios, thelowest being about 2/I while the average is about 8/1. Ol all associatedolivines reported in the literature, the highest Fe/Mg ratio is If 2 andmostare lower. It was by similar interpretation of data that Ross and Shannon(1926) were led to state, "the proportion of silica has remained nearlyconstant . . . , the iron has all been changed from the ferrous to the ferricstate and its proportion has greatly increased, water has been added inlarge amount, and magnesium has been largely abstracted. It is clearthat in the change of olivine to iddingsite there has been a metasomaticreplacement ," h addition Ross and Shannon, noting the occurrenceof fresh olivine rims surrounding altered olivine cores and fresh olivinemicrolites with altered olivine phenocrysts, state that alteration is partlydependent on composition of the olivine. Although these relations indi-cate selective replacement of magnesian olivine (assuming normal zon-ing), some authors (e.g., Tomkeiefi, 1934) have interpreted this ambilu-ous statement to mean selective alteration of fayalitic olivine. Typicallythe initial stages of alteration are characterized not by zonal replacementof olivine, but by replacement along fractures randomly cutting thephenocrysts. Since textural relationships indicate either no zonal prefer-ence or preference for magnesian zones, composition differences betweenolivine and its alteration products are best explained by Ross' and Shan-non's conclusion that I4g is leached. The common occurrence of iron ox-ides with the clay minerals and the not rare complete replacement ofolivine by iron oxides indicates addition of iron as well as removal ofmagnesium. It is of course clear that the alteration cannot be simply aprocess of hydration and oxidation, for the proportions of FeO (recalcu-lated from Fe2Oa), MgO and SiO: do not approach those of olivine. It isnot, therefore, unreasonable to suppose that compositions of relicts arerepresentative of the entire original phenocryst subject to zoning fluctua-tions no more severe than found in unaltered olivine.

These changes in bulk composition and structure have resulted inalteratiori products which are considerably Iess dense (see Table 4A')than the minerals they replace. Since the alteration is pseudomorphic,volume relations require removal of material, which, according to theabove arguments, involves selective leaching of Mg. Ross and Shannondo not consider the ultimate disposition of the extracted material, butPeacock and Fuller (1928) clearly state their belief that chlorophaeite

ALTERATION OF OLIVINE AND ORTHOPYROXENE

Tasr,E 4A. Csrlrrcal Alq.qr-vsrs*

t41

A B C

Average Range Average Range Average Range

SiOzTiOrAlzo:FesOaFeOMnOMsoCaONuroKzOHzO+

47.20 22.0H8.3(i0 .14 0 .72 - 0 .2 ( ;3.81 0.00- 5.0! l

35.42 25.48-d0.1:i0 .38 0 .00 - 1 .970 .05 0 .00 - 0 .116.85 0 .60-12 .712 .35 1 . 72 - . s .2 ( t0 .12 0 .12 - 0 . l : i0 .10 0 .09 - 0 .109 .30 5 .74 -11 .00

44.79 42.6846.320.22 0 .00- 0 .468.24 5 .87-70.628.90 6 .33-12.304 . 7 7 2 . 7 0 , 7 . 9 8

2 2 . 5 0 2 1 . 7 1 - 2 3 . 2 97 . 9 6 0 . 3 2 - 3 . 2 9

u.84 7 .47-10.00

45.11 43 .4248.770.29 0 .00- 0 .823 . 9 1 7 . 3 7 - 7 . 1 2

22.13 12.37-30 026 . 6 6 2 . s 6 , r 2 . 1 00 . 4 7 0 . 3 5 - 0 . 6 28 16 6.26-11 733 . 1 3 1 . 9 7 - 4 . 0 20.55 0 0a- 2 .210 . 1 4 t r - 0 . 4 78.39 6 .83-10.97

Total 99.54 100.22 98.94

H:O- 9 .84 4.76-22 .8: i 10 .72 7 .69-13.52 20.78 12.5+-27.44

Den-sity 2 . 5 - 2 . 8 2 . 2 4 - 2 . 3 0 1 . 8 1 - 2 . 2 3

* ColumnA (iddingsite) represents 10 analyses, 2 from Tomkeiefi (1934), 8 from Rcssand Shannon (1926). Column B (saponite) represents 4 analyses from Caillere and Henin(1951). Column C (chlorophaeitr:) represents 4 analyses from Peacock and Fuller (1928).

All analyses were recalculated to l007o excluding HzO- before averages were com-puted.

vesicle-filling and veinlets result from an alteration of mafic minerals. Inthe same manner, it is believed that material extracted from mafic min-erals during alteration to a,nisotropic varieties is recombined in the samestructural types (principally smectite-chlorite) in joints and vesicles.

The limited chemical deuta presented in Table 4 may now be dividedinto two groups:1) those representing extracted material in motion atthe time of consolidation ticolumns B and C, Table 4A; specimens 6, 8,and 10, Table 48), and2) those representing alteration residues (columnA, Table 4A). It is clear by comparison of these groups that, in basaltIavas, both are characterirzed by high Fe/Mg ratios. It is reasonable toask, then, why pseudomol:phs are retained at all if the material whichprecipitates from the final solutions is oI the same composition as that inpseudomorphs. This is probably largely due to concentration of iron inthe altering solutions; as the mafic minerals break down, l\tlg passes intosolution over a concentration gradient while iron is recombined in placein a solid phase approprial.e to the environment. lt is presumed that Mg

H. G. WILSHIRE

Taslr 4B. Cnnurcnr, Axer.Ysas

g* 4^' loL/

SiOzA10sFezo;l'eOMgoCaOTiOz

12 .800 . 4 36 . 6 7

2 5 . 2 4

nd

7 . 1 40 . 8 06 . 0 9

3 1 . 8 0

o . 3 7

49.+85 . 8 5

18 .863 .90

12.662 . 7 30 . 3 5

Rock Analyses{

Unaltered Altered' , . ' - -

gms/+1' Unaltered gms/ry$ Altered

SiOzAlzOgFezOrFeOMgo

CaONurOKsOTiOz

COzHrO+

140.043 .02 . 1

26 .s1 9 921 .9

t . o+ . 74 . 30 . 01 . 8

134. 544.414.412 .7l J . ,

a a n

2 . 74 . O0 . 0o . 5

5 1 . 6 6 5 1 . 5 11 5 . 8 8 1 6 . 9 80 . 7 6 5 . 4 99 . 7 0 4 . 8 27 . 3 5 5 . 2 68 . 0 7 8 . 5 62 . 8 1 2 . 1 71 . 5 2 1 . 0 51 . 6 0 1 . 6 8nd nd0. 65 2 .48

Total

x Numbers correspond to same samples in other tables. See Table 1 for descriptions.

t Olivine basalt near Orange, New South Wales. See text for description. Analyses

recalculated to 100/s excluding HzO-.Analysts: Avery and Anderson, Melbourne.

added to the solutions is outweighed by Fe, the final crystall ine productstill being Fe-rich. Since the structural groups represented by the altera-tion products can accommodate any variation in Fe, Mg and Al, the pres-

ence of these minerals in joints and vesicles does not imply alteration ofmafic minerals, but the twg usually occur together. When this material isaccommodated interstitially, its composition can have only an indirectbearing on the nature of liquid residues formed by fractional crystalliza-tion.

The data on alteration products from shallow intrusions indicate thatthe material in motion is rich in NIg and relatively poor in Fe. Since

ALTERATION OII OLIVINE AND ORTHOPYROXENE I43

associated Fe-bearing minr:rals are rare in joint-filling, this is taken toindicate a lack of iron enrichment in the residual solutions. Since neitherMg nor AI can be expected to be concentrated in the altering solutions,complete solution of mafic minerals susceptible to the alteration is possi-ble. Where pseudomorphs are retained, duration of alteration may bemore important than comp,osition of solutions, but additional factors ofstructural inheritance in a,lteration products and heat supply provideproblems for which there are no quantitative data. The markedly higheralumina may be attributed to contributions from altered Al-bearing min-erals such as plagioclase which is much more extensively altered in theserocks than in typical alterr:d basalt lavas.

Itlncunwrslrs ol nlovnlrENT oF Exrnecrsr Marnnrar,

Since there is good eviclence that alteration takes place, at least insome lava flows, at a late stage of consolidation, some mechanism forrapid removal of material t:xtracted during alteration is required. Occur-rence of alteration producl-s as delicate networks of veinlets connectingpseudomorphs and vesicles, as veinlets parallel to flow banding, and,principally, along major joints all suggest operation of a pressure gradi-ent. Smedes and Lang (1955) appealed to jointing to provide low pressureareas towards which alter:ing solutions moved. If alteration precedesjointing, extracted material could be removed in this manner. Polygonalnetworks of veinlets may be interpreted as micro-joints and may haveserved the important function of cleaning out on a small scale. Veinletsparallel to flow banding could be caused by shearing along irregular flowplanes resulting in dilated areas of low pressure. Curtis (1954) utilizedthe last proposal in interpreting autobrecciation of lavas. Whatever themechanism of cleaning out. on a small scale, the major joint system actsas a permanent low pressure area until pressures are equdlized, and it istowards these that the altering fluids tend to move. In shallow intru-sions, low pressure areas are probably created chiefly by jointing, thelocation of much of the smectite-chlorite, but this wil l depend greatlyon the nature of the wall rock.

Errncrs oF THE ArrBt.qrroN oN BuLK Cnpmrc.q.r Corvrposrrror.tor Soun Bas,lrrrc Rocrs

It is clear from the con,clusions drawn above that alteration of maficminerals results in selective leaching of NIg which, in lavas, is later com-bined with iron and silica prresumably derived from the altering solutions.If these components are not recombined within the rock, but escape intojoints or wall rock, the co:mposition of the rock no longer represents thecomposition of the magma from which it crystallized. The quantitativeefiects of the alteration can be assessed only where fresh and altered

144 H. G, WILSHIRE

equivalents occur. The effects of loss of Fe, silica and other componentsconcentrated in solution prior to alteration cannot be quantitativelyassessed.

In Table 48 are given chemical analyses of altered and unaltered por-tions of the Orange, New South Wales basalt. These differ only in nearlycomplete alteration of olivine and minor alteration of glass in the alteredphase. Since density changes have taken place, comparison of weightper cents does not give a true picture of composition changes and thesehave been recalculated to gms./m3. Specimens were collected from severalcolumns and combined to reduce sampling error. The most significantchanges are loss of N{gO and SiO2, addition of HzO and change of ferrousto ferric iron in the altered rock. T'otal iron and alkalis show a small lossand alumina a small gain in the altered phase, but, in view of samplingand analytical error, these cannot be safely interpreted. The analysis ofsmectite-chlorite joint-filling from this locality (no. 6, Table 48) clearlyindicates the final disposition of Mg leached from olivine. It also showsthat the Fe content of the fresh rock is not representative of the magma.

Data presented by Fuller (1938) indicate that opposite changes in bulkcomposition can take place under certain conditions. In the basalt flowsdescribed by Fuller, alteration of olivine is confined to cores of columns,residues of the alteration being accommodated interstitially. Althoughdirect comparison of analyses shows litt le change between altered coresand unaltered sheaths, Fuller (Pers. comm.) believes that the poroustexture of the sheaths is explained by loss of the mesostasis (formed as acrystallization residue, not by alteration). Under these conditions recalcu-lation of weight per cent analyses to gms./m.3 would probably showchanges in the opposite direction from those described above.

It is necessary to bear in mind that these minerals are only a smallpart of a much more complex association. This is particularly true ofshallow intrusions where carbonates and zeolites are invariable associatesof the clay minerals and the total secondary mineral assemblage ex-tremely varied. Since these too are usually less dense than the primaryminerals from which they are derived, their escape into joints may be ex-pected to have a similar effect on bulk chemical composition of the alteredrocks.

ApprrcerroN To SoME Pnonr,nlrs rN SorL Cr,av MrNBnALocy

It is apparent from the results of this investigation that chlorites andsmectites occurring in soils derived from basic volcanic rocks cannot beassumed to have originated by weathering. The possibility that theseminerals remain as relicts, although probably modified depending onconditions of weathering, is strengthened by a detailed study of the

ALTERATION O1' OLIV INE AND ORTHOPYROXENE 145

Carmel Bay, California basrllts which lie in a temperate climate zone, and

which are structurally disp,osed to allow through drainage. A section is

accessible at this locality which contains rocks ranging from those whose

only altered constituent is olivine to those so deeply weathered

that all of the primary minerals are altered and the rock is very porous.

The freshest rocks contain green alteration products which are triocta-

hedral smectite-chlorite ag;gregates, while the most highly weathered

rocks contain oxidized smectite-chlorite and goethite aggregates, the

smectite-chlorite of these rocks being predominantly intermediate to

dioctahedral. By plotting the position of the 060 reflection against aqualitative estimate of degr:ee of weathering of the enclosing rock it was

found that the higher the degree of weathering the more closely the

smectite-chlorite approach,ed a dioctahedral structure. X-ray powderphotographs of a colorless clay pseudomorphing clinopyroxene-plagio-clase aggregates (presumed to have formed by weathering because suchalteration occurs only in the weathered zone) show that it is also com-posed in part of smectite-chlorite aggregates, but difiers in having kao-

Iinite (?) as part of the aggregate. It might be expected, since the struc-

tural types represented by deu'teric alteration products are stable in the

weathering environment while pyroxene and plagioclase undergo a com-plete structural change, that conversion of these to dioctahedral struc-

tures will lag behind alteration of plagioclase and pyroxene. This was

found to be true since psr:udomorphs after olivine, even in the most

deeply weathered rocks, are still composed in part of a trioctahedralstructure. In addition, tht: occurrence of lightly pigmented (goethite)

smectite-chlorite in deeply weathered, almost colorless rocks attests to

continued leaching of irorL. Although weathering of these rocks tookplace in a temperate climate with through drainage, it is obvious thatweathering conditions whi,:h favor formation of trioctahedral chlorites

and smectites will also fal'or preservation of the original trioctahedralproducts of deuteric alteration. It should be noted, however, that weath-

ering of smectite-chlorite in the Orange, New South Wales basalt resulted

in an aggregate of goethite, qtartz and kaolinite without preservation of

the smectite-chlorite.It is believed that recognition of the abundance and character of these

minerals as pre-weathering constituents of basic igneous rocks will aid in

solution of some problems in soil-clay mineralogy which have heretofore

been explained by compli,:ated chemical or physical histories such as

those described by Allen and Scheid (1946) and Ross and Hendricks(1945). In no case wil l i t b,e a simple problem to evaluate residual con-

tributions to the clay frat:tion, for these minerals are not evenly dis-

tributed through the parerLt rock. This is particularly true of differenti-

146 I]. G. WILSHIRE

ated shallow intrusions such as that at Prospect, New South Wales,where the smectite-chlorite content of difierent specimens ranges frompractically nil to more than half by volume,

AcrNowrplcMENr

The author is indebted to Drs. C. D. Campbell, A. B. Edwards, B. C.King, R. L. Rose, A. C. Waters, L. Weiss and H. Will iams for specimensused in the study. \{embers of the Geology and Physics Departments,Univ. of 'fechnology,

Sydney, are gratefully acknowledged for makingequipment available for the author's use. Special thanks are due Profs.I{owel Williams and Charles l{eyer of the University of California forhaving crit ically read the manuscript, and to Drs. R. E. Fuller and G. F.Walker for their helpful discussion of some of the problems. ProfessorPabst of the University of California kindly supplied powder photographsof iron oxides for comparison. Financial assistance was received from theUniversity Research Grants Committee, Sydney University.

RnrrnnNcps

Ar.r.nN, V. T., aNn Scnnro, V.E. (1946), NontroniJe in the Columbia River region: Am.M i.ner atr., 31, 2941372.

Bnr.nr.rv, W. F. (1945), Glycol-montmorillonite : A m. Mineral., 30, 704-7 13.Bnrxornv, G. W. (ed.) (1951), X-ray identification and crystal structures of clay minerals:

The Mineralogical Society, London.BnowN, G. (1955), Report of the clay minerals group sub-committee on nomenclature of

clay minerals: Clay Mins. 8u11.,2,296.Cerr,rona, S. (1935), Sur les caracteres specifiques de la bowlingite: Compt. Rend. Acail.

Sci. Paris, 200, 1483-1485.Catlnnr, S. exo HnNrN, S. (1951), The properties and identification of saponite (bowling-

ite): Clay Mins. Bull.,l, 138-144.Coers, R. R. (1940), Propylitization and related types of alteration on the Comstock Lode,

Nevada: Econ. Geol.,35, 1-16.Cuntrs, G. H. (1954), Mode of origin of pyroclastic debris in the Mehrten Formation of

the Sierra Nevada: Unit:. oJ CaIiJ. Pub. in Geol. 9ci.,29,453-502.Eanlnv, J. W. nm Mrr.Nn, I. H. (1956), Regularly interstratified montmorillonite-

chlorite in basalt: Proc. zlth Nat'I. Conf . on Clays and CIay Mi.ns., pub.456, 381-384.Eowrnns, A. B. (1938), The formation of iddingsite: Am. Mineral.,23,277-281.Furlrn, R. E. (1938), Deuteric alteration controlled by the jointing of lavas: Am. Jour.

Sci.,35, 161-171.Hewnev, J. B. (1877), On bowlingite, a new Scottish Mineral: Min. Mog.,l, 154t-157.Hlncnnavns, A., eNo Tavron, W. H. (1946), An r-ray examination of decomposition

products of chrysotile (asbestos) and serpentine: Min. Mag.,27,204-216.IonrNcs, J. P. (1892), U. S. Geol. Suntey., Mono. 20, Append. B, 388-390.KalnnN, G. vrw orn (1951), Optical studies on natural plagioclase feldspars with high-

and low-temperature optics: Pub. lan het Min. Geol,. Inst. R1fts., Univ., Utrecht,Holland.

Lawsow, A. C. (1893), The geology of Carmelo Bay Unit. of CaliJ. Pub. in GeoL Sci., l,1-59.

ALTERATION OP OLIVINE AND ORTHOPVROXENE l4i

MtNc-SnaN SuN (1957), The nature of iddingsite in some basaltic rocks of New Mexico:

Am. M ina aL, 42, 525-533.

Pnecocr, M. A., eNo Fur.ren, IL. E. (1928), Chlorophaeite, sideromelane and palagonite

from the Columbia River Plateau: Am. Mineral., t3' 360-382.

Pnrnen, R. T., aNo Cor-r, W. F. (1942), The alteration products of olivine and leucite in

the leucitelamproites from the west Kimberly area, western Australia: Am. Mineral.,

27,373-384.Ross, C. S., aNo Horonrcrs, S. B. (1945), Minerals of the montmorillonite group: their

origin and relation to soils and clays: U. S. Geol. Sura. Prof . Po'per no.2O5-8r 23-79'

Ross, C. S., eNn SnmwoN, E. V, (1926), The origin, occurrence, composition and physical

properties of the mineral idclingsite: Proc. IJ. S. Nat'|. Museutn,67, Art.7,l-19'

Sltnons, H. W., aNo Lnxc, A. J., Jn. (1955), Basalt column rinds caused by deuteric

alteration: Am. Iour. Sci.,253, 173*781.

TourrmFr, S. I. (1934), Differentiation in basalt lava, Island Magee, Co. Antim: Geol'

Mag.,71,501-512.Tnorcnn, W. E. (1952), Tabr:llen zur optischen Bestimmung der gesteinsbildenden

Minerale : E. Schweizerbartr;che Verlagsbuchhandlung, Stuttgart.

Tunxrn, F. J. Q947), Determination of plagioclase with the four-axis universal stage:

A m. M i,na aI., 32, 389-410.

Manuscribt recehd. June 4. I95r'.